Machine learning approaches for imaging-based prognostication of the outcome of surgery for mesial temporal lobe epilepsy.

OBJECTIVES: Around 30% of patients undergoing surgical resection for drug-resistant mesial temporal lobe epilepsy (MTLE) do not obtain seizure freedom. Success of anterior temporal lobe resection (ATLR) critically depends on the careful selection of surgical candidates, aiming at optimizing seizure freedom while minimizing postoperative morbidity. Structural MRI and FDG-PET neuroimaging are routinely used in presurgical assessment and guide the decision to proceed to surgery. In this study, we evaluate the potential of machine learning techniques applied to standard presurgical MRI and PET imaging features to provide enhanced prognostic value relative to current practice. METHODS: Eighty two patients with drug resistant MTLE were scanned with FDG-PET pre-surgery and T1-weighted MRI pre- and postsurgery. From these images the following features of interest were derived: volume of temporal lobe (TL) hypometabolism, % of extratemporal hypometabolism, presence of contralateral TL hypometabolism, presence of hippocampal sclerosis, laterality of seizure onset volume of tissue resected and % of temporal lobe hypometabolism resected. These measures were used as predictor variables in logistic regression, support vector machines, random forests and artificial neural networks. RESULTS: In the study cohort, 24 of 82 (28.3%) who underwent an ATLR for drug-resistant MTLE did not achieve Engel Class I (i.e., free of disabling seizures) outcome at a minimum of 2 years of postoperative follow-up. We found that machine learning approaches were able to predict up to 73% of the 24 ATLR surgical patients who did not achieve a Class I outcome, at the expense of incorrect prediction for up to 31% of patients who did achieve a Class I outcome. Overall accuracies ranged from 70% to 80%, with an area under the receiver operating characteristic curve (AUC) of .75-.81. We additionally found that information regarding overall extent of both total and significantly hypometabolic tissue resected was crucial to predictive performance, with AUC dropping to .59-.62 using presurgical information alone. Incorporating the laterality of seizure onset and the choice of machine learning algorithm did not significantly change predictive performance. SIGNIFICANCE: Collectively, these results indicate that "acceptable" to "good" patient-specific prognostication for drug-resistant MTLE surgery is feasible with machine learning approaches utilizing commonly collected imaging modalities, but that information on the surgical resection region is critical for optimal prognostication.

Outcomes of extended transforaminal lumbar interbody fusion for lumbar spondylosis.

This study aims to assess the results of extended transforaminal lumbar interbody fusion (TLIF) for a two surgeon, single institution series. In total, extended TLIF with bilateral decompression was performed in 57 patients. Pain, American Spinal Injury Association scores, patient demographics, body mass index (BMI), perioperative indices and radiographic measurements were recorded and analysed. The surgeries were performed between February 2011 and January 2014 on 38 women and 19 men. The mean patient age was 62.86 years, and the mean BMI was 30.31 kg/m(2). In 49 patients, spondylolisthesis was the primary indication. The mean intraoperative time was 284.65 min, and this decreased as the series progressed. The median length of stay was 5 days (range: 2-9). The surgical complication rate was 19.3%. Two patients died from cardiopulmonary complications. Single level TLIF was performed in 78.9% of the cohort, with L4/5 the most commonly fused level. Significant pain reduction was achieved from a mean (± standard deviation) preoperative visual analogue scale (VAS) of 8.28 ± 1.39 to 1.50 ± 1.05 at 12 months postoperatively. No patients deteriorated neurologically. Spondylolisthesis was significantly corrected from a preoperative mean of 6.82 mm to 2.80 mm postoperatively. Although there is a learning curve associated with the procedure, extended TLIF with bilateral facet joint removal and decompression appeared to be a safe and effective alternative to other fusion techniques, and our results were comparable to other published case series. The stabilisation and correction of spinal deformity reduces pain, aids neurologic recovery and improves quality of life.

Visual outcomes and headache following interventions for idiopathic intracranial hypertension.

The optimal surgical management for medically refractory idiopathic intracranial hypertension (IIH) is not well established. Few studies have directly compared headache and visual outcomes across treatment modalities. A systematic analysis of case series was conducted to compare therapeutic efficacies among currently available interventions. The electronic databases from EMBASE (1980-17 September 2013), Medline (1980-17 September 2013), Cochrane databases, and references of review articles was searched. All publications reporting headache and visual outcomes following intervention for IIH were included. A total of 457 manuscripts were selected and full text analysis produced 30 studies with extractable data. All studies constituted Class III evidence. Overall, 332 patients treated by optic nerve sheath fenestration (ONSF), 287 by lumboperitoneal shunt (LPS), 61 by ventriculoperitoneal shunt (VPS), and 88 by dural venous sinus stenting, were identified. Visual acuity improved in 49.3%, 56.6%, 67.2% and 84.6% of patients following VPS, LPS, ONSF, and stent placements, respectively. Resolution of papilledema was noted in 59.9% to 97.1%. Postoperative headache improved in 36.5%, 62.5%, 75.2%, and 82.9% of patients treated with ONSF, VPS, LPS, and stenting, respectively. Shunt revision was more frequent for LPS compared to VPS (46% versus 36%; p<0.2). Among the LPS revisions, 87.5% occurred within the first 12 months following initial surgery. Our pooled analysis indicated an overall similar improvement in visual outcomes across treatment modalities, and a modest improvement in headache following cerebrospinal fluid shunting and endovascular stent placement. Based on currently available literature, there is insufficient evidence to recommend or reject any treatments modalities for IIH.

Management of recurrent Cushing's disease.

The management of patients with Cushing's disease remains a major challenge. The remission rate for initial transsphenoidal surgery is reported as 69-94% while the recurrence rate following initially successful transsphenoidal surgery is 2-27%. Treatment options for Cushing's disease after failed initial therapy or recurrence include repeat transsphenoidal surgery, radiotherapy, medical therapy and bilateral adrenalectomy. Management options following unsuccessful initial surgery or recurrence are presented and discussed.

Tailored cortical resection following image guided subdural grid implantation for medically refractory epilepsy.

The aim of this study was to report the safety and efficacy of tailored cortical resection based on image guided subdural electrode implantations in eight patients with medically refractory epilepsy. The patients were selected for multimodality image guided subdural grid implantation, inpatient invasive electroencephalography video monitoring and surgical resection of epileptogenic foci. All patients had frequent disabling, medically refractory seizures pre-operatively. At a minimum of 10 months post-resection all patients had a worthwhile improvement in seizure frequency, with 7 of the 8 (87.5%) having an excellent outcome (Engel Class I). Short-term complications of grid implantation were: one patient with a post-operative subdural haemorrhage and one patient with a transient fluctuating dysphasia. The only long-term complication was a mild, non-disabling dysarthria following resection near eloquent speech cortex in one patient. We conclude that tailored cortical resection following image-guided insertion of subdural grids is a reliable, safe and highly effective method for the treatment of medically refractory epilepsy in carefully selected patients.