Search Results (78)

Perioptic meningiomas pose a therapeutic challenge due to their proximity to critical visual structures. Single-fraction stereotactic radiosurgery is known to effectively control the growth of meningiomas, but this subgroup carries the risk of optic neuropathy, which is minimized with the introduction of dose hypofractionation. The Leksell Gamma Knife Icon has perfected fractionated stereotactic radiosurgery, maintaining submillimeter accuracy in each dose fraction without the need for an invasive frame. This study analyzes the feasibility, safety, and efficacy of multi-fraction Gamma Knife Icon radiosurgery for perioptic meningiomas, taking into account tumor control rates, visual preservation, and treatment-related toxicity. We conducted a retrospective analysis of 100 patients with a perioptic meningioma treated with fractionated Gamma Knife Icon radiosurgery between September 2017 and December 2022. A total of 80 Patients were female, and 20 were male; the mean age was 61.7 years (range 35–84). The most frequent anatomical locations included: cavernous sinus (35 pts), anterior clinoid (17 pts), sphenoid wing (14 pts) and olfactory groove (11 pts). The median tumor volume was 5.6 mL (range 0.12–31.7 mL). Most patients (89%) received 25 Gy in five fractions. Tumor control was achieved in 98% of cases, with a mean radiological follow-up of 41.2 months. Tumor volume did not predict radiological shrinkage (p = 0.639). Tumor shrinkage was observed more frequently in the no prior surgery group (p = 0.035). The mean clinical follow-up was 45.3 months. Among symptomatic patients (35 pts) at baseline, symptoms remained stable in 27 (77%) cases, improved in 5 (14%), and worsened in 3 (9%). No new symptoms were observed in asymptomatic patients. Overall clinical deterioration occurred in three (3%) patients—one because of tumor progression; although, without statistical evidence (p = 0.217), worsened patients had notable larger mean tumor volumes (12.6 mL vs. 6.8 mL). The dosimetric advantages of Gamma Knife technology are empowered by the biological benefits of fractionation and the convenience of non-invasive immobilization. Excellent tumor control rates and positive visual outcomes favor its routine application in properly selected patients. Full article

Background: Surgical resection remains the standard treatment for large brain metastases (LBMs), but many patients are not surgical candidates due to poor performance status or uncontrolled systemic disease. Gamma Knife-based hypofractionated stereotactic radiotherapy (GKRS) has emerged as a potential alternative; however, its clinical role in this population remains insufficiently defined. We evaluated whether a uniform daily 5-fraction GKRS provides effective and safe local treatment for surgery-ineligible LBMs. Methods: We retrospectively analyzed 100 patients with LBMs (>14 cm³) who underwent primary hypofractionated GKRS using a uniform daily 5-fraction schedule. Forty-six patients were male; the median age was 60 years. The median Karnofsky Performance Status (KPS) was 70 (60–100); a total of 47 patients (47%) had pre-GKRS neurological deficits. The most common primary sites were lung (41), breast (24), and kidney (14). The median tumor volume was 22.0 cm³ (14–70 cm³), and the marginal dose was 35.2 Gy (50% isodose line) in 5 fractions. The primary endpoints included local tumor control (LTC), intracranial progression-free survival (PFS), and overall survival (OS). Radiation necrosis (RN) was assessed as a key safety outcome. Results: At a median follow-up of 18 months, the overall LTC rate was 74%, with 1-, 2-, and 3-year rates of 73%, 65%, and 60%, respectively. Median PFS and OS were 7.5 and 16.3 months. Higher pre-treatment KPS and absence of neurological deficits were independently associated with improved OS (p = 0.003 and 0.025, respectively). RN occurred in 16% of patients, with 9% developing symptoms; all symptomatic cases were effectively managed with corticosteroids or bevacizumab. Most tumors demonstrated substantial volumetric reduction, with a median decrease of 80% and 30% achieving near-complete response (>95%). Conclusions: A uniform daily 5-fraction hypofractionated GKRS provides effective local control with acceptable toxicity in patients with LBMs. These findings support its role as a feasible local treatment option in selected patients who are not candidates for surgery. Integration with systemic therapies and prospective validation are warranted to refine patient selection and optimize outcomes. Full article

Background. Patients with malignant or benign central nervous system (CNS) tumours are evaluated for suitability of treatment modality based on multiple clinical and tumour-related factors. To obtain multidisciplinary consensus, a patient’s file and imaging are commonly reviewed by a tumour board (TB). There are three relevant weekly TB venues at our institute—gamma knife stereotactic radiosurgery (SRS) intake rounds, CNS rounds, and stereotactic body radiotherapy (SBRT) rounds—which are attended by non-overlapping clinician teams. We explored the clinical parameters prompting multiple TB reviews in patients with complex CNS tumours. Methods. Data were retrospectively obtained from electronic medical records. Patients referred for discussion at SRS rounds (November 2017–June 2020) were cross-referenced with those reviewed in CNS rounds and SBRT rounds. The cohort of interest included patients who underwent review at more than one TB for the same indication. Patient, tumour, and treatment factors were abstracted, and descriptive statistics were calculated. A sub-cohort of patients with pre-plans created for both SRS and conventionally fractionated external beam radiotherapy (EBRT) was identified. Dosimetric data were analyzed. Results. Of 1091 patients, 87 (8.0%) were discussed at more than one TB. 59/87 (67.8%) patients were reviewed at two TBs pertaining to the same CNS lesion and comprised the study cohort. The most common tumour type was meningioma (20/59), and the most common reason for multiple discussions was proximity to optic structures (19/59). After TB discussions, 25/59 patients were seen in consultation by one specialist, 29/59 by two, and 5/59 by none. Overall, the final treatment decisions were conventional EBRT in 21/59; SRS in 18/59; surveillance in 12/59; surgery in 3/59; systemic therapy in 3/59; proton referral in 1/59; and SBRT in 1/59. A total of 20/59 patients were treated with palliative intent. Among all patients who ultimately received radiotherapy, median interval between the first TB discussion and the first RT treatment was 56 days (IQR 7.5–65.5 d). The pre-plan sub-cohort consisted of four patients, all of whom were ultimately treated with conventional EBRT. Conclusions. Evidence to support optimal treatment for some complex CNS tumours can be limited. Multiple radiotherapy modalities may be equally favourable (or unfavourable) options. Proximity to the optic apparatus and previous CNS irradiation are common reasons for clinical equipoise. Tumour board review is an essential tool in formulating a multidisciplinary care plan; however, attention should be paid to ensuring that subsequent consultations and treatment initiation are not unduly delayed. Full article

Background/Objectives: LINAC-based single-isocenter (SIT) stereotactic radiosurgery (SRS) enables efficient treatment of multiple brain metastases but may compromise target conformity and increase low-dose brain exposure, particularly for spatially distributed lesions. Dual-isocenter techniques (DITs) may mitigate these limitations, while Gamma Knife (GK) remains the reference standard for high-selectivity radiosurgery. This study compares SIT- and DIT LINAC-based SRS with GK, focusing on target conformity and low-dose brain exposure under equivalent, zero-margin targeting assumptions. Methods: Twenty-eight patients with multiple brain metastases (197 lesions) were included in this retrospective planning study. For each patient, three plans were generated: a GK plan and LINAC-based SIT and DIT plans using automated dynamic conformal arc optimization (Elements Multiple Brain Metastases). All plans were generated using a zero-millimeter GTV-to-PTV margin strategy. For DIT, lesions were automatically clustered and assigned to two isocenters. Target coverage required ≥99% of each GTV to receive the prescription dose. Plan quality was evaluated using the Paddick Conformity Index (PCI) on a per-lesion basis and low-dose brain volumes (V12, V10, V5, V4, and V3 Gy) on a per-patient basis. Paired non-parametric tests and multivariable models were used to assess technique-related differences and associations with total target volume and lesion count. Results: GK achieved the highest median PCI (0.83), followed closely by DIT (0.77), while SIT plans demonstrated significantly lower conformity (0.73). Compared with GK, the median PCI difference was −0.05 for DIT and −0.08 for SIT. Conformity for DIT remained stable across lesion volumes and lesion counts, whereas GK conformity increased modestly with lesion size. Low-dose brain exposure differed significantly between techniques at all dose levels (p < 0.001). GK consistently yielded the lowest Vx volumes, SIT the highest, and DIT intermediate values. Relative to GK, SIT plans showed progressively larger increases in low-dose exposure at lower dose levels (mean ΔV3 ≈ +149 cc), while DIT reduced this low-dose spread (mean ΔV3 ≈ +117 cc). Total target volume was the dominant predictor of low-dose brain exposure across all techniques, with a smaller additional contribution from lesion count. Conclusions: DIT LINAC-based SRS significantly improves target conformity and reduces low-dose brain exposure compared with SIT delivery, achieving dosimetric performance that closely approximates Gamma Knife under equivalent zero-margin targeting assumptions. While Gamma Knife remains the reference standard for low-dose sparing, dual-isocenter planning represents a clinically robust and scalable alternative that effectively balances plan quality and treatment efficiency in patients with multiple brain metastases. Full article

Background/Objectives: Our study aims to identify potential new MRI features of brain metastases (BMs) that could be further used in overall survival (OS) assessment. Methods: A total of 109 patients with BMs were included. Kaplan–Meier analysis, the log-rank test, and Cox Regression were implemented in the survival analysis. The first ten significant features were incorporated into four distinct machine learning (ML) algorithms to predict six-month survival. Results: Survival analysis revealed that multiple brain lesions and synchronous presentation were associated with a poor prognostic value (HR > 1; p = 0.01, p = 0.02). Other features demonstrated a protective effect on OS including the absence of extracranial lesions (HR < 1, p = 0.04) and the presence of solid enhancement (HR < 1, p < 0.05). In this observational cohort, treatment was associated with longer OS—including surgery, gamma knife radiosurgery, whole brain radiation therapy, and chemotherapy—compared to best supportive care (HR < 1, p < 0.005); these treatment-related hazard ratios are not interpreted causally. The shallow Neural Networks model was the top-performing ML model, achieving an AUC of 0.93 (CI = 0.89–0.97). According to the Shapley Additive Explanations analysis, the solid enhancement type had a positive impact on OS, whereas a higher number of lesions, larger volumes and a cystic morphology were associated with negative outcomes. Conclusions: Our results confirm that including morphological MRI features of BMs in the prediction of OS significantly contributes to the enhancement of ML algorithms’ prediction and discriminatory capacity. Full article

Background: Gamma Knife radiosurgery (GKS) delivers highly conformal intracranial irradiation, yet clinical decision-making still relies predominantly on physical dose metrics that do not account for fractionation, dose rate, treatment time, or DNA repair. Classical radiobiological models—including the linear–quadratic (LQ) formula and the Jones–Hopewell single-session repair model—do not extend naturally to 3- and 5-fraction GKS. Meanwhile, growing evidence suggests that biologically effective dose (BED) may better capture radiosurgical response in selected pathologies. A unified, biologically grounded, multi-fraction GKS framework has been lacking. Methods: We developed AI-BED-Fx, the first multi-fraction extension of the Jones–Hopewell radiobiological model capable of computing fraction-resolved BED for 1-, 3-, and 5-fraction GKS. The framework incorporates α/β ratio, dual-component repair kinetics, isocentre geometry, beam-on–time structure, and lesion-specific biological parameters. Four synthetic pathology-specific cohorts—arteriovenous malformation (AVM), meningioma (MEN), vestibular schwannoma (VS), and brain metastasis (BM)—were generated using distinct radiobiological signatures. Machine-learning models were trained to quantify the predictive value of physical dose versus BED for local control or obliteration. Additional experiments included Bayesian estimation of α/β and a neural-network surrogate for fast BED prediction. An exploratory comparison with a 60-lesion clinical brain–metastasis dataset was performed to assess whether key trends observed in the synthetic BM cohort were consistent with real radiosurgical outcomes. Results: AI-BED-Fx produced realistic pathology-specific BED distributions (AVM 60–210 Gy_2.47; MEN 41–85 Gy_3.5; VS 46–68 Gy₃; BM 37–75 Gy₁₀) and biologically coherent dose–response relationships. Predictive modeling demonstrated strong pathology dependence. In AVM, the three models achieved AUCs of 0.921 (Model A), 0.922 (Model B), and 0.924 (Model C), with corresponding Brier scores of 0.054, 0.051, and 0.051, with BED-based models performing best. In meningioma, BED was the dominant predictor, with AUCs of 0.642 (Model A), 0.660 (Model B), and 0.661 (Model C) and Brier scores of 0.181, 0.177, and 0.179, respectively. In vestibular schwannoma, the narrow BED range resulted in minimal BED contribution, with AUCs of 0.812, 0.827, and 0.830 and Brier scores of 0.165, 0.160, and 0.162, with physical dose and tumor volume determining performance. In brain metastases, outcomes were driven primarily by volume and physical dose, with AUCs of 0.614, 0.630, and 0.629 and Brier scores of 0.254, 0.250, and 0.253, showing negligible improvement from BED. AI-BED-Fx also accurately recovered the true α/β from synthetic outcomes (posterior mean 2.54 vs. true 2.47), and a neural-network surrogate reproduced full radiobiological BED calculations with near-perfect fidelity (R² = 0.9991). Conclusions: AI-BED-Fx provides the first unified, biologically explicit framework for modeling single- and multi-fraction Gamma Knife radiosurgery. The findings show that the predictive usefulness of BED is pathology-specific rather than universal, and that radiobiological dose provides additional predictive value only when repair kinetics and dose–response biology support it. By integrating mechanistic radiobiology with machine learning, AI-BED-Fx establishes the conceptual and computational foundations for biologically adaptive, AI-guided radiosurgery, and cross-pathology comparison of treatment response. This work uses large radiobiologically grounded synthetic cohorts for methodological validation; limited real-patient data are included only for exploratory consistency checks, and full clinical validation is planned. Full article

Background/Objectives: Hemangioblastomas are rare, benign, highly vascular tumors of the central nervous system, frequently associated with von Hippel–Lindau (vHL) disease. Case Presentation: We report a 16-year-old female with vHL presenting with recurrent headaches, abdominal distension, and ocular discomfort. Imaging revealed hemangioblastomas in the fourth ventricle and retrobulbar space, alongside multiple pancreatic cysts. The patient underwent three sessions of Gamma Knife Surgery (GKS) with initial tumor regression and symptom relief. However, long-term follow-up demonstrated progressive disease, with new lesions in the cerebellum, spinal cord, and orbit, including cystic transformation. Histopathology confirmed the reticular variant of hemangioblastoma. Despite further radiosurgical and surgical recommendations, the patient and family opted for conservative management, with lesions remaining radiographically stable over nine years. Conclusions: This case demonstrates that Gamma Knife Surgery may provide temporary local disease control for selected solid hemangioblastomas in von Hippel–Lindau disease but does not alter the underlying disease course. Long-term radiographic stability should be interpreted cautiously, as hemangioblastomas exhibit saltatory growth patterns that make it difficult to distinguish treatment effect from natural tumor quiescence. These findings emphasize that radiosurgery should be regarded as a disease-control strategy rather than curative therapy, underscoring the importance of individualized management, multidisciplinary decision-making, and prolonged surveillance. Full article

Primary intracranial squamous cell carcinoma (SCC) arising from an epidermoid cyst is an exceptionally rare and aggressive malignancy with a dismal prognosis. Conventional management typically involves gross total resection followed by wide-field radiotherapy; however, this intensive approach is often unfeasible for elderly or frail patients. We present a case of primary intracranial SCC in a 75-year-old woman who presented with rapid cochleovestibular deterioration. Imaging revealed subtle enlargement of a long-standing cerebellopontine angle epidermoid cyst. Subtotal resection was performed to preserve critical neurovascular structures adherent to the infiltrative tumor. Given the patient’s poor performance status and the risk of toxicity from broad-field radiation, adjuvant Gamma Knife radiosurgery (GKS) was selected as a focal salvage modality. Despite the limited surgical margin, the patient has maintained a progression-free status with no radiographic evidence of disease progression for 18 months without neurological decline. This case highlights the diagnostic challenge of malignant transformation disguised by radiologic mimicry and demonstrates that GKS can serve as an effective and tolerable adjuvant strategy. We propose that for high-risk patients precluded from intensive multimodal therapy, focused stereotactic irradiation offers a viable alternative to secure local control while preserving quality of life and systemic immune integrity. Full article

Background/Objectives: Hypofractionated stereotactic radiosurgery (hfSRS) is increasingly used for benign intracranial tumors that are large or located near critical neural structures to reduce treatment-related toxicity. However, the optimal interval between fractions remains poorly defined, particularly for slowly proliferating benign tumors. This study evaluated clinical outcomes and longitudinal volumetric response patterns following Gamma Knife hfSRS delivered at fixed two-week intervals, with particular attention to the biological relevance of fraction timing. Methods: We retrospectively analyzed 126 patients with benign intracranial tumors, including meningioma, non-functioning pituitary neuroendocrine tumor (PitNET), vestibular schwannoma, and craniopharyngioma, treated between 2016 and 2022. Treatment was delivered in 2–5 fractions at fixed two-week intervals using Gamma Knife radiosurgery. Radiological outcomes included tumor control rate and longitudinal volumetric changes, while clinical outcomes included visual, auditory, and endocrine function. Propensity score matching was performed in PitNET and vestibular schwannoma cohorts to compare hfSRS with single-fraction stereotactic radiosurgery while minimizing baseline imbalances. Results: The overall tumor control rate was 98.4%. Across the entire cohort, tumors demonstrated a median volume reduction of −0.64% per month. In the propensity score–matched PitNET cohort, tumor control was comparable between treatment groups, whereas hfSRS was associated with earlier and greater volumetric reduction over time compared with single-fraction treatment. In the matched vestibular schwannoma cohort, long-term tumor control was similar between groups; however, transient tumor enlargement occurred more frequently after hfSRS without adversely affecting long-term tumor control or functional hearing outcomes. Conclusions: Hypofractionated stereotactic radiosurgery delivered at fixed two-week intervals achieved excellent tumor control with acceptable toxicity in selected benign intracranial tumors. These findings support the clinical feasibility of a fixed two-week inter-fraction interval and suggest that fraction timing may represent a biologically relevant treatment parameter influencing early volumetric response patterns without compromising long-term outcomes. Full article

Epilepsy remains an active and important area of research due to its complex etiology, significant global burden, and variable response to treatment. Current knowledge has provided valuable insights into the underlying molecular mechanisms of the disease and continues to guide the development of novel therapeutic strategies. This review presents a comprehensive overview of the etiologies of epilepsy, as well as traditional and modern medical and surgical treatment approaches, while highlighting future research directions. Peer-reviewed articles retrieved from PubMed and Google Scholar were analyzed and synthesized to produce this review. The etiological complexity of epilepsy arises from genetic, metabolic, structural, and inflammatory mechanisms, which often coexist rather than act independently. A wide range of anti-seizure drugs (ASDs) is currently available, with many new agents targeting novel mechanisms under development. Surgical approaches, including resection, disconnection, corpus callosotomy, and neuromodulation, are widely used for patients with drug-resistant epilepsy and result in variable seizure outcomes. In addition, minimally invasive techniques such as laser interstitial thermal therapy (LITT), stereoelectroencephalography-guided radiofrequency thermocoagulation, gamma knife radiosurgery, and high-intensity focused ultrasound have gained clinical relevance and continue to be explored. Emerging technologies, including artificial intelligence, machine learning, and precision medicine, offer promising directions for future research. Although several potential biomarkers have been identified, none are yet established for routine clinical use. Continued investigation is essential to improve understanding of epileptogenesis and to develop safer, more effective therapies. Full article

Background: Gamma Knife radiosurgery (GKS) is widely used for the management of intracranial disorders. Emerging evidence suggests that incorporating the biological effective dose (BED) into GKS planning may improve the prediction of treatment efficacy and toxicity. This review aims to evaluate the role of BED in GKS across multiple intracranial indications. Methods: A qualitative review of published clinical studies was performed to assess the application of BED models in GKS for pituitary adenomas, vestibular schwannomas, meningiomas, arteriovenous malformations (AVMs), trigeminal neuralgia, and other disorders. The relationships between BED, treatment outcomes, and adverse effects were compared across indications. Results: The association between BED and clinical outcomes was most consistent in AVMs, where higher BED correlated closely with obliteration rates. In other diseases, BED-based analyses showed promising but variable predictive value. Notably, BED-derived parameters demonstrated improved prediction of post-GKS hypopituitarism in pituitary adenomas and AVM obliteration compared with physical dose alone. However, most available evidence was derived from retrospective studies. Conclusions: BED may serve as a valuable complement to conventional physical dose metrics in GKS planning, but its ability to replace physical dose remains uncertain. Prospective studies and histology-specific radiobiological parameter validation are required to establish the routine clinical utility of BED. Full article

Background/Objectives: Posterior fossa brain metastases (PFBMs) pose particular risks owing to their proximity to the brainstem and fourth ventricle. We evaluated the safety (treatment-related complications), local effectiveness, and procedural efficiency of volumetric modulated arc therapy (VMAT)-based stereotactic radiosurgery (VMAT-SRS) for PFBMs. Methods: This single-centre, retrospective study derived a PFBM subgroup from an overall institutional cohort of 123 patients treated with VMAT-RapidArc SRS/fSRS. The doses were 12–20 Gy (single fraction) or 5 × 6 Gy (selected cases). Local response (mRECIST) and predefined safety endpoints (symptomatic oedema with brainstem/IV-ventricle compromise, obstructive hydrocephalus, haemorrhagic transformation, CSF diversion, and urgent neurosurgery) were assessed. Overall survival and procedural time were analysed. Results: Thirty-one patients (39 lesions) were included; 76.9% of them received single-fraction SRS. In addition, 74.2% of patients had supratentorial metastases with posterior fossa involvement. Kaplan–Meier overall survival at 6, 12, 24, and 48 months was 74%, 58%, 26%, and 9.7%, respectively; the median survival time was 12.6 months. Among evaluable lesions, local control was 84.5% (per-lesion response: 15.5% PD, 28.1% SD, 34.4% PR, and 22.0% CR). No clinically significant posterior fossa local complications were observed. Three patients developed radiation-induced leukoencephalopathy after whole-brain radiotherapy (WBRT) and radiosurgery for synchronous supratentorial metastases. The median procedural time was 25.0 min (IQR 9.0) with one isocentre versus 52.5 min (IQR 9.75) with two. Conclusions: VMAT-SRS/fSRS for PFBMs achieved high local control, very low posterior fossa toxicity, and favourable procedural efficiency, supporting its use as a safe, rapid, frameless alternative to WBRT and other radiosurgical platforms such as Gamma Knife in appropriately selected patients. Full article

Background and Clinical Significance: Meningiomas are among the most common primary intracranial tumors, usually benign and slow-growing. Extracranial extension is exceptionally rare, particularly when arising from convexity meningiomas extending into the temporal fossa. Such cases pose unique diagnostic and therapeutic challenges due to their atypical growth patterns and anatomical complexity. Case Presentation: A 63-year-old woman previously treated for a right temporal convexity meningioma with subtotal resection and Gamma Knife radiosurgery demonstrated progressive extracranial tumor growth over five years, while the intracranial component remained stable. MRI revealed infiltration of the temporalis and lateral pterygoid muscles and erosion of the temporal bone. Due to extensive extracranial involvement and limited neurosurgical accessibility, resection was performed by a maxillofacial surgical team through a preauricular approach. Intraoperatively, the tumor was encapsulated but adherent to the deep temporal fascia and zygomatic arch. The temporal branch of the facial nerve was identified and preserved. Histopathology confirmed a meningothelial meningioma, WHO Grade I, with low proliferative activity (Ki-67 < 1%). Postoperative recovery was uneventful, with transient facial nerve weakness that resolved within weeks. Conclusions: This report adds to the limited literature describing temporal fossa involvement by convexity meningiomas and illustrates the value of collaboration between neurosurgical and maxillofacial teams. Regular MRI surveillance every 6–12 months is advised for early detection of recurrence. Full article

Background: Radiation therapy is a key treatment modality for brain metastases. While providing a treatment alternative, post-treatment imaging often presents diagnostic challenges, particularly in distinguishing tumor recurrence from radiation-induced changes such as necrosis. Advanced imaging techniques and artificial intelligence (AI)-based radiomic analyses emerge as alternatives to help lesion characterization. The objective of this study was to assess the capacity of machine learning algorithms to distinguish between brain metastases recurrence and radiation necrosis. Methods: The research was conducted in two phases and used publicly available MRI data from patients treated with Gamma Knife radiosurgery. In the first phase, 30 cases of local recurrence of brain metastases and 30 cases of radiation-induced necrosis were considered. Image segmentation and radiomic feature extraction were performed on these data using MatRadiomics_1_5_3, a MATLAB-based framework integrating PyRadiomics. Features were then selected using point-biserial correlation. In the second phase, a classification was performed using a Support Vector Machine model with repeated stratified cross-validation settings. Results: The results achieved an accuracy on the test set of 83% for distinguishing metastases from necrosis. Conclusions: The results of this feasibility study demonstrate the potential of radiomics and AI to improve diagnostic accuracy and personalized care in neuro-oncology. Full article

Background/Objectives: GK plan quality is strongly affected by lesion size and shape, and the same evaluation metrics may not be directly comparable across patients with different anatomies. This study proposes a deep learning-based method to predict achievable, clinically acceptable plan quality from patient-specific geometry. Methods: A hierarchically densely connected U-Net (HD-U-Net) was trained at the lesion level to predict 3D dose distributions for the estimation of plan quality metrics, including coverage, selectivity, gradient index (GI), and conformity index at a 50% prescription dose (CI50). To improve the prediction accuracy of plan quality metrics, Dice similarity coefficient losses for the 100% and 50% isodose lines were incorporated with conventional mean squared error (MSE) loss. Results: Ten-fold cross-validation on 463 brain metastases (BMs) from 175 patients showed that our method achieved smaller mean absolute errors across all four metrics than the HD-U-Net baseline trained with MSE loss. Improvements were pronounced in all metrics for small metastases, and were observed primarily in GI and CI50 for medium and large lesions. Paired Wilcoxon signed-rank tests confirmed the statistical significance of these improvements (p < 0.05). Conclusions: The proposed method outperformed the baseline model in capturing overall trends, improving per-lesion accuracy, and enhancing robustness to dataset variability. It can serve as a pre-planning tool to guide planners in constraint setting and priority tuning, a post-planning quality control tool to identify subpar plans that could be substantially improved, and as a foundation for developing deep reinforcement learning-based automated planning of GK treatments for brain metastases. Full article