RESUMO
Glioblastoma (IDH-wildtype) represents a formidable challenge in oncology, lacking effective chemotherapeutic or biological interventions. The metabolic reprogramming of cancer cells is a hallmark of tumor progression and drug resistance, yet the role of metabolic reprogramming in glioblastoma during drug treatment remains poorly understood. The dihydroorotate dehydrogenase (DHODH) inhibitor BAY2402234 is a blood-brain barrier penetrant drug showing efficiency in in vivo models of many brain cancers. In this study, we investigated the effect of BAY2402234 in regulating the metabolic phenotype of EGFRWT and EGFRvIII patient-derived glioblastoma cell lines. Our findings reveal the selective cytotoxicity of BAY2402234 toward EGFRWT glioblastoma subtypes with minimal effect on EGFRvIII patient cells. At sublethal doses, BAY2402234 induces triglyceride synthesis at the expense of membrane lipid synthesis and fatty acid oxidation in EGFRWT glioblastoma cells, while these effects are not observed in EGFRvIII glioblastoma cells. Furthermore, BAY2402234 reduced the abundance of signaling lipid species in EGFRWT glioblastoma. This study elucidates genetic mutation-specific metabolic plasticity and efficacy in glioblastoma cells in response to drug treatment, offering insights into therapeutic avenues for precision medicine approaches.
RESUMO
Gliomas are the most prevalent malignancies of the central nervous system (CNS). Downregulation of microRNA124 (miR124) has been identified in glioma; however, its biological functions in glioma are not yet fully understood. Specificity protein 1 (SP1) is a type of transcription factor that is involved in cancer progression. In this study, we examined the targeting of Sp1 mRNA by miR-124-3p in a rat glioma model. After confirming and selecting the binding of Sp1 to miR-124 with the help of bioinformatics methods, adult male Wistar rats were used to induce glioma by microinjection of 1 × 106 C6 cells into the striatum area of brain. The rats were divided into 3 groups; intact, sham and glioma groups. The presence of glioma was confirmed 21 days after implantation through histological analysis. The expression levels of miR-124 and SP1 genes in the experimental groups were examined using quantitative real-time polymerase chain reaction (qRT-PCR). Our data showed that the expression of miR-124 was significantly downregulated in glioma group compared to the sham and intact group, while the expression of SP1 was significantly upregulated. We found that the expression levels of miR-124 and Sp1 were decreased and increased in C6 cell line compared to the normal brain tissue cell line, respectively. The results indicated that Sp1 was identified as a direct target of miR124 through luciferase reporter assays. In summary, this study demonstrated for the first time that miR-124 expression is downregulated and Sp1 expression is upregulated in an animal model of glioma, which, in turn, may be involved in the development of glioma brain cancer.
Assuntos
Neoplasias Encefálicas , Regulação Neoplásica da Expressão Gênica , Glioma , MicroRNAs , Ratos Wistar , Fator de Transcrição Sp1 , Animais , MicroRNAs/genética , MicroRNAs/metabolismo , Glioma/genética , Glioma/patologia , Glioma/metabolismo , Fator de Transcrição Sp1/metabolismo , Fator de Transcrição Sp1/genética , Masculino , Ratos , Linhagem Celular Tumoral , Neoplasias Encefálicas/genética , Neoplasias Encefálicas/patologia , Neoplasias Encefálicas/metabolismo , Progressão da Doença , Regulação para BaixoRESUMO
Background and purpose: The brainstem tumour known as diffuse intrinsic pontine glioma (DIPG), also known as pontine glioma, infiltrative brainstem glioma is uncommon and virtually always affects children. A pontine glioma develops in the brainstem's most vulnerable region (the "pons"), which regulates a number of vital processes like respiration and blood pressure. It is particularly challenging to treat due to its location and how it invades healthy brain tissue. The hunt for a solution is continually advancing thanks to advances in modern medicine, but the correct approach is still elusive. With a particular focus on brain tumours that are incurable or recur, research is ongoing to discover fresh, practical approaches to target particular areas of the brain. Experimental approach: To successfully complete this task, a thorough literature search was carried out in reputable databases like Google Scholar, PubMed, and ScienceDirect. Key results: The present article provides a comprehensive analysis of the notable advantages of lipid nanoparticles compared to alternative nanoparticle formulations. The article delves into the intricate realm of diverse lipid-based nanoparticulate delivery systems, which are used in Diffuse Intrinsic Pontine Glioma (DIPG) which thoroughly examines preclinical and clinical studies, providing a comprehensive analysis of the effectiveness and potential of lipid nanoparticles in driving therapeutic advancements for DIPG. Conclusion: There is strong clinical data to support the promising method of using lipid-based nanoparticulate drug delivery for brain cancer treatment, which shows improved outcomes.
RESUMO
Patients affected by glioma frequently suffer of epileptic discharges, however the causes of brain tumor-related epilepsy (BTRE) are still not completely understood. We investigated the mechanisms underlying BTRE by analyzing the effects of exosomes released by U87 glioma cells and by patient-derived glioma cells. Rat hippocampal neurons incubated for 24 h with these exosomes exhibited increased spontaneous firing, while their resting membrane potential shifted positively by 10-15 mV. Voltage clamp recordings demonstrated that the activation of the Na+ current shifted towards more hyperpolarized voltages by 10-15 mV. To understand the factors inducing hyperexcitability we focused on exosomal cytokines. Western Blot and ELISA assays show that TNF-α is present inside glioma-derived exosomes. Remarkably, incubation with TNF-α fully mimicked the phenotype induced by exosomes, with neurons firing continuously, while their resting membrane potential shifted positively. RT-PCR revealed that both exosomes and TNF-α induced over-expression of the voltage-gated Na channel Nav1.6, a low-threshold Na+ channel responsible for hyperexcitability. When neurons were preincubated with Infliximab, a specific TNF-α inhibitor, the hyperexcitability induced by exosomes and TNF-α were drastically reduced. We propose that Infliximab, an FDA approved drug to treat rheumatoid arthritis, could ameliorate the conditions of glioma patients suffering of BTRE.
RESUMO
Emerging Industry 5.0 designs promote artificial intelligence services and data-driven applications across multiple places with varying ownership that need special data protection and privacy considerations to prevent the disclosure of private information to outsiders. Due to this, federated learning offers a method for improving machine-learning models without accessing the train data at a single manufacturing facility. We provide a self-adaptive framework for federated machine learning of healthcare intelligent systems in this research. Our method takes into account the participating parties at various levels of healthcare ecosystem abstraction. Each hospital trains its local model internally in a self-adaptive style and transmits it to the centralized server for universal model optimization and communication cycle reduction. To represent a multi-task optimization issue, we split the dataset into as many subsets as devices. Each device selects the most advantageous subset for every local iteration of the model. On a training dataset, our initial study demonstrates the algorithm's ability to converge various hospital and device counts. By merging a federated machine-learning approach with advanced deep machine-learning models, we can simply and accurately predict multidisciplinary cancer diseases in the human body. Furthermore, in the smart healthcare industry 5.0, the results of federated machine learning approaches are used to validate multidisciplinary cancer disease prediction. The proposed adaptive federated machine learning methodology achieved 90.0%, while the conventional federated learning approach achieved 87.30%, both of which were higher than the previous state-of-the-art methodologies for cancer disease prediction in the smart healthcare industry 5.0.
Assuntos
Aprendizado de Máquina , Neoplasias , Humanos , Setor de Assistência à Saúde , Algoritmos , Inteligência Artificial , Atenção à SaúdeRESUMO
Soloxolone amides are semisynthetic triterpenoids that can cross the blood-brain barrier and inhibit glioblastoma growth both in vitro and in vivo. Here we investigate the impact of these compounds on processes associated with glioblastoma invasiveness and therapy resistance. Screening of soloxolone amides against glioblastoma cells revealed the ability of compound 7 (soloxolone para-methylanilide) to inhibit transforming growth factor-beta 1 (TGF-ß1)-induced glial-mesenchymal transition Compound 7 inhibited morphological changes, wound healing, transwell migration, and expression of mesenchymal markers (N-cadherin, fibronectin, Slug) in TGF-ß1-induced U87 and U118 glioblastoma cells, while restoring their adhesiveness. Confocal microscopy and molecular docking showed that 7 reduced SMAD2/3 nuclear translocation probably by direct interaction with the TGF-ß type I and type II receptors (TßRI/II). In addition, 7 suppressed stemness of glioblastoma cells as evidenced by inhibition of colony forming ability, spheroid growth, and aldehyde dehydrogenase (ALDH) activity. Furthermore, 7 exhibited a synergistic effect with temozolomide (TMZ) on glioblastoma cell viability. Using N-acetyl-L-cysteine (NAC) and flow cytometry analysis of Annexin V-FITC-, propidium iodide-, and DCFDA-stained cells, 7 was found to synergize the cytotoxicity of TMZ by inducing ROS-dependent apoptosis. Further in vivo studies showed that 7, alone or in combination with TMZ, effectively suppressed the growth of U87 xenograft tumors in mice. Thus, 7 demonstrated promising potential as a component of combination therapy for glioblastoma, reducing its invasiveness and increasing its sensitivity to chemotherapy.
RESUMO
Despite being the leading cause of childhood mortality, pediatric gliomas have been relatively understudied, and the repurposing of immunotherapies has not been successful. Whole transcriptome sequencing, single-cell sequencing, and sequential multiplex immunofluorescence were used to identify an immunotherapy strategy evaluated in multiple preclinical glioma models. MAPK-driven pediatric gliomas have a higher interferon signature relative to other molecular subgroups. Single-cell sequencing identified an activated and cytotoxic microglia population designated MG-Act in BRAF-fused MAPK-activated pilocytic astrocytoma (PA), but not in high-grade gliomas or normal brain. TIM3 is expressed on MG-Act and on the myeloid cells lining the tumor vasculature but not normal brain. TIM3 expression becomes upregulated on immune cells in the PA microenvironment and anti-TIM3 reprograms ex vivo immune cells from human PAs to a pro-inflammatory cytotoxic phenotype. In a genetically engineered murine model of MAPK-driven low-grade gliomas, anti-TIM3 treatment increased median survival over IgG and anti-PD1 treated mice. ScRNA sequencing data during the therapeutic window of anti-TIM3 demonstrates enrichment of the MG-Act population. The therapeutic activity of anti-TIM3 is abrogated in the CX3CR1 microglia knockout background. These data support the use of anti-TIM3 in clinical trials of pediatric low-grade MAPK-driven gliomas.
RESUMO
Brain cancer is one of the deadliest diseases, although many efforts have been made to treat it, there is no comprehensive and effective treatment approach yet. In recent years, the use of network-based analysis to identify important biological genes and pathways involved in various complex diseases, including brain cancer, has attracted the attention of researchers. The goal of this manuscript is to perform a comprehensive analysis of the various results presented related to brain cancer. For this purpose, firstly, based on the CORMINE medical database, collected all the genes related to brain cancer with a valid P-value. Then the structural and functional relationships between the above gene sets have been identified based on the STRING database. Next, in the PPI network, hub centrality analysis was performed to determine the proteins that have many connections with other proteins. After the modularization of the network, the module with the most hub vertices is considered as the most relevant module to the formation and progression of brain cancer. Since the driver vertices play an important role in biological systems, the edges of the selected module were oriented, and by analyzing the controllability of complex networks, a set of five proteins with the highest control power has been identified. Finally, based on the drug-gene interaction, a set of drugs effective on each of the driver genes has been obtained, which can potentially be used as new combination drugs. Validation of the hub and driver proteins shows that they are mainly essential proteins in the biological processes related to the various cancers and therefore the drugs that affect them can be considered as new combination therapy. The presented procedure can be used for any other complex disease.
Assuntos
Neoplasias Encefálicas , Redes Reguladoras de Genes , Mapas de Interação de Proteínas , Humanos , Neoplasias Encefálicas/genética , Neoplasias Encefálicas/tratamento farmacológico , Mapas de Interação de Proteínas/genética , Mapas de Interação de Proteínas/efeitos dos fármacos , Biologia Computacional/métodos , Regulação Neoplásica da Expressão Gênica/efeitos dos fármacos , Antineoplásicos/uso terapêutico , Protocolos de Quimioterapia Combinada Antineoplásica/uso terapêuticoRESUMO
Tumor-educated platelets (TEPs) are a potential method of liquid biopsy for the diagnosis and monitoring of cancer. However, the mechanism underlying tumor education of platelets is not known, and transcripts associated with TEPs are often not tumor-associated transcripts. We demonstrated that direct tumor transfer of transcripts to circulating platelets is an unlikely source of the TEP signal. We used CDSeq, a latent Dirichlet allocation algorithm, to deconvolute the TEP signal in blood samples from patients with glioblastoma. We demonstrated that a substantial proportion of transcripts in the platelet transcriptome are derived from non-platelet cells, and the use of this algorithm allows the removal of contaminant transcripts. Furthermore, we used the results of this algorithm to demonstrate that TEPs represent a subset of more activated platelets, which also contain transcripts normally associated with non-platelet inflammatory cells, suggesting that these inflammatory cells, possibly in the tumor microenvironment, transfer transcripts to platelets that are then found in circulation. Our analysis suggests a useful and efficient method of processing TEP transcriptomic data to enable the isolation of a unique TEP signal associated with specific tumors.
RESUMO
Quiescence is a prolonged but reversible state of cell-cycle arrest that is an adaptive feature of most adult stem cell populations. In the brain, quiescence helps to protect adult neural stem cells from stress and supports lifelong neurogenesis. Unfortunately however, entry into a quiescent or a slow-cycling state is also a malignant feature of brain cancer stem cells. In glioblastoma, where the process has been best characterised, quiescent glioma stem cells preferentially survive chemoradiation, and after therapy, reactivate to regrow the tumour and drive recurrence. In this Review, we discuss the in vitro and in vivo models that have been developed for studying neural stem cell quiescence and how these tools may be used to deepen biological understanding and to develop novel therapies targeting quiescent glioma stem cells.
RESUMO
Ayurveda is the traditional medicine system of India, and has been in practice for millennia. It is a traditional approach that uses 1000's of different plant preparations in various combinations for treatment of human ailments, including cancer. Ethnopharmacological and phytochemical analyses are now elucidating the bioactive constituents of the different plant species and herbal formulations, including ashwagandha, curcumin, guduchi, triphala, and others. To provide an overview of: 1) the ethnopharmacology of Ayurveda and several of its most important plant species and formulations, including pharmacological and molecular mechanisms of its anti-cancer effects; 2) review the literature applying Ayurvedic herbs and formulations to brain tumors. A detailed PubMed search was performed that included publications involving Ayurveda, cancer, ethnopharmacology, phytochemical analysis, molecular analysis, and brain tumors. In recent decades, significant research has begun to elucidate the bioactive compounds of ashwagandha, tumeric, guduchi, and triphala, such as withaferin A, withanolides, curcumin, palmatine, and many others. These compounds and extracts are now being applied to brain tumor cells in vitro and in animal models, with positive signs of anti-cancer activity including reduced cell growth, increased apoptosis, cell cycle arrest, increased differentiation, and inhibition of important internal signal transduction pathways. Several Ayurvedic herbs (ashwagandha, curcumin) have bioactive compounds with significant anti-cancer activity, and are effective in early pre-clinical testing against brain tumor cells in vitro and in animal models. Further pre-clinical testing is warranted, along with advancement into phase I and phase II clinical trials of patients with glioblastoma and other brain tumors.
RESUMO
The mostly aggressive and extremely malignant type of central nervous system is Glioblastoma (GBM), which is characterized by an extremely short average survival time of lesser than 16 months. The primary cause of this phenomenon can be attributed to the extensively altered genome of GBM, which is characterized by the dysregulation of numerous critical signaling pathways and epigenetics regulations associated with proliferation, cellular growth, survival, and apoptosis. In light of this, different genetic alterations in critical signaling pathways and various epigenetics regulation mechanisms are associated with GBM and identified as distinguishing markers. Such GBM prognostic alterations are identified in PI3K/AKT, p53, RTK, RAS, RB, STAT3 and ZIP4 signaling pathways, metabolic pathway (IDH1/2), as well as alterations in epigenetic regulation genes (MGMT, CDKN2A-p16INK4aCDKN2B-p15INK4b). The exploration of innovative diagnostic and therapeutic approaches that specifically target these pathways is utmost importance to enhance the future medication for GBM. This study provides a comprehensive overview of dysregulated epigenetic mechanisms and signaling pathways due to mutations, methylation, and copy number alterations of in critical genes in GBM with prevalence and emphasizing their significance.
RESUMO
The article presents, for the first time, a terahertz metamaterial absorber (TMA) designed in the shape of a cross consisting of four orthogonally positioned horn-shaped patches in succession, to detect brain cancer cells. The design exhibits the property of mu-negative material, indicating magnetic resonance. The proposed TMA has achieved an impressive absorption rate of 99.43% at 2.334 THz and a high Q-factor of 47.15. The sensing capability has been investigated by altering the refractive index of the surrounding medium in the range of 1.3 to 1.48, resulting in a sensitivity of 0.502 THz/RIU. The proposed TMA exhibits complete polarization insensitivity, highlighting this as one of its advantageous features. The adequate sensing capability of the proposed TMA in differentiating normal and cancerous brain cells makes it a viable candidate for an early and efficient brain cancer detector. This research can be the foundation for future research on using THz radiation for brain cancer detection.
Assuntos
Neoplasias Encefálicas , Radiação Terahertz , Neoplasias Encefálicas/diagnóstico por imagem , Humanos , Refratometria/métodos , Encéfalo/diagnóstico por imagem , Desenho de EquipamentoRESUMO
Pediatric diffuse midline gliomas (DMG) with H3-K27M-altered are aggressive brain tumors that arise during childhood. Despite advances in genomic knowledge and the significant number of clinical trials testing new targeted therapies, patient outcomes are still insufficient. Immune checkpoint blockades with small molecules, such as aptamers, are opening new therapeutic options that represent hope for this orphan disease. Here, we demonstrated that a TIM-3 aptamer as monotherapy increased the immune infiltration and elicited a strong specific immune response with a tendency to improve the overall survival of treated DMG-bearing mice. Importantly, combining TIM-3 Apt with radiotherapy increased the overall median survival and led to long-term survivor mice in two pediatric DMG orthotopic murine models. Interestingly, TIM-3 aptamer administration increased the number of myeloid populations and the pro-inflammatory ratios of CD8: Tregs in the tumor microenvironment as compared to non-treated groups after radiotherapy. Importantly, the depletion of T-cells led to a major loss of the therapeutic effect achieved by the combination. This work uncovers TIM-3 targeting as an immunotherapy approach to improve the radiotherapy outcome in DMGs and offers a strong foundation for propelling a phase I clinical trial using radiotherapy and TIM-3 blockade combination as a treatment for these tumors.
RESUMO
Background: Glioblastoma, a high-grade primary brain cancer, has a median survival of approximately 14 months. Post-mortem brain donation provides insight to pathogenesis along with spatial and temporal heterogeneity. Post-mortem brain biobanking programs are increasing in number and the need to understand and improve the associated human experience is pressing. This study aims to qualitatively explore the experiences of next of kin (NOK) following the death and brain donation of a loved one and to understand the impact such programs have on NOK carers. Method: We interviewed 29 NOK following the death of their loved one and subsequent brain donation. Thematic analysis was conducted on the transcribed, qualitative interviews. Results: Four themes were identified; (1) Brain donation is a straightforward decision grounded in altruism and pragmatism; (2) Supporting donors is a source of comfort, pride and empowerment; (3) Brain donation can provide meaning for suffering and tragedy and (4) Perceptions of procedures and processes when supporting a loved one to donate. Insights into areas for improvement, for example transporting donors following a home death and the role of the body bag were also noted. Conclusion: Supporting a loved one to donate their brain can be a positive experience providing a source of hope, empowerment and purpose for NOK. Data indicating areas for consideration are broadly relevant for improving the delivery of brain donation programs for future donors and their loved ones.
Understanding how loved ones feel about someone close to them donating their brain to research after their death from brain cancer The act of donating brain tissue after death from brain cancer is a huge gift to medical research and may have an impact on the ability of the scientific community to improve outcomes for people diagnosed with brain cancers. While we understand how valuable these donations are for research, we need more work to understand how these donations impact the people who donate and those who love and support them. This paper explores the experiences of people who have lost someone to brain cancer who then went on to donate their brain tissue after their death. Through the use of interviews, it explores the impact that the donation has on a loved one or next of kin from providing a source of comfort, empowerment, pride or an alternative to 'senseless' suffering and tragedy. It also provides areas that should be considered by people who are facilitating brain donations to ensure that any potential, harm or upset can be minimized.
RESUMO
PURPOSE: Patients with brain cancer and painful symptoms of the disease experience heavy pressure and negative inner experiences, leading to a sense of stigma. Therefore, this study assessed the level of stigma in patients with brain cancer and analyzed the risk factors for stigma to analyze the underlying relationships among depression, social support, low self-esteem, and stigma. METHODS: Patients completed the Social Impact Scale, Self-rating Depression Scale, Rosenberg Self-Esteem Scale, Herth Hope Index, Social Support Rating Scale, and Self-Perceived Burden Scale. Multiple linear regression analysis was used to identify factors independently associated with stigma. Parallel mediation analysis was used to evaluate the mediating role of the relationship between psychoemotional factors and stigma. RESULTS: A multivariate linear regression analysis demonstrated significant associations between age (ß = - 0.189, P = 0.002), treatment (ß = 0.184, P = 0.003), self-esteem (ß = - 0.128, P = 0.046), depression (ß = 0.273, P < 0.001), hope (ß = - 0.217, P = 0.003), and self-perceived burden (ß = 0.260, P < 0.001) with brain cancer. It was observed that the social support received by brain cancer patients directly impacted their stigma (total effect, - 0.851, P = 0.001). Additionally, this relationship was influenced by depression and self-esteem through two distinct pathways. CONCLUSION: Increased stigma among brain cancer patients was found to be associated with severe depression, feelings of inferiority, diminished hope, and a heavy perceived burden. The structural equation modeling (SEM) revealed that social support negatively influenced stigma through depression and self-esteem. It is imperative to grasp patients' inner needs, implement psychological interventions, and cultivate a cancer-friendly social environment to prevent stigmatization and discrimination based on their patient status.
Assuntos
Neoplasias Encefálicas , Depressão , Análise de Mediação , Autoimagem , Estigma Social , Apoio Social , Humanos , Estudos Transversais , Masculino , Feminino , Pessoa de Meia-Idade , Neoplasias Encefálicas/psicologia , Depressão/psicologia , Depressão/etiologia , Adulto , Idoso , Fatores de Risco , Escalas de Graduação Psiquiátrica , Inquéritos e Questionários , EsperançaRESUMO
The fight against cancer, a relentless global health crisis, emphasizes the urgency for efficient and automated early detection methods. To address this critical need, this review assesses recent advances in non-invasive cancer prediction techniques, comparing conventional machine learning (CML) and deep neural networks (DNNs). Focusing on these seven major cancers, we analyze 310 publications spanning the years 2018 to 2024, focusing on detection accuracy as the key metric to identify the most effective predictive models, highlighting critical gaps in current methodologies, and suggesting directions for future research. We further delved into factors like datasets, features, and modalities to gain a comprehensive understanding of each approach's performance. Separate review tables for each cancer type and approach facilitated comparisons between top performers (accuracy exceeding 99%) and low performers (65.83 to 85.8%). Our exploration of public databases and commonly used classifiers revealed that optimal combinations of features, datasets, and models can achieve up to 100% accuracy for both CML and DNN. However, significant variations in accuracy (up to 35%) were observed, particularly when optimization was lacking. Notably, colorectal cancer exhibited the lowest accuracy (DNN 69%, CML 65.83%). A five-point comparative analysis (best/worst models, performance gap, average accuracy, and research trends) revealed that while DNN research is gaining momentum, CML approaches remain competitive, even outperforming DNN in some cases. This study presents an in-depth comparative analysis of CML and DNN techniques for cancer detection. This knowledge can inform future research directions and contribute to the development of increasingly accurate and reliable cancer detection tools.
RESUMO
Brain tumors have a poor prognosis. Early, accurate diagnosis and treatment are crucial. Although brain surgical biopsy can provide an accurate diagnosis, it is highly invasive and risky and is not suitable for follow-up examination. Blood-based liquid biopsies have a low detection rate of tumor biomarkers and limited evaluation ability due to the existence of the blood-brain barrier (BBB). The BBB is composed of brain capillary endothelial cells through tight junctions, which prevents the release of brain tumor markers to the human peripheral circulation, making it more difficult to diagnose, predict prognosis, and evaluate therapeutic response through brain tumor markers than other tumors. Focused ultrasound (FUS)-enabled liquid biopsy (sonobiopsy) is an emerging technique using FUS to promote the release of tumor markers into the circulatory system and cerebrospinal fluid, thus facilitating tumor detection. The feasibility and safety data from both animal models and clinical trials support sonobiopsy as a great potential in the diagnosis of brain diseases.
RESUMO
Gamma Knife radiosurgery (GKRS) is a well-established technique in radiation therapy (RT) for treating small-size brain tumors. It administers highly concentrated doses during each treatment fraction, with even minor dose errors posing a significant risk of causing severe damage to healthy tissues. It underscores the critical need for precise and meticulous precision in GKRS. However, the planning process for GKRS is complex and time-consuming, heavily reliant on the expertise of medical physicists. Incorporating deep learning approaches for GKRS dose prediction can reduce this dependency, improve planning efficiency and homogeneity, streamline clinical workflows, and reduce patient lagging times. Despite this, precise Gamma Knife plan dose distribution prediction using existing models remains a significant challenge. The complexity stems from the intricate nature of dose distributions, subtle contrasts in CT scans, and the interdependence of dosimetric metrics. To overcome these challenges, we have developed a "Cascaded-Deep-Supervised" Convolutional Neural Network (CDS-CNN) that employs a hybrid-weighted optimization scheme. Our innovative method incorporates multi-level deep supervision and a strategic sequential multi-network training approach. It enables the extraction of intra-slice and inter-slice features, leading to more realistic dose predictions with additional contextual information. CDS-CNN was trained and evaluated using data from 105 brain cancer patients who underwent GKRS treatment, with 85 cases used for training and 20 for testing. Quantitative assessments and statistical analyses demonstrated high consistency between the predicted dose distributions and the reference doses from the treatment planning system (TPS). The 3D overall gamma passing rates (GPRs) reached 97.15% ± 1.36% (3 mm/3%, 10% threshold), surpassing the previous best performance by 2.53% using the 3D Dense U-Net model. When evaluated against more stringent criteria (2 mm/3%, 10% threshold, and 1 mm/3%, 10% threshold), the overall GPRs still achieved 96.53% ± 1.08% and 95.03% ± 1.18%. Furthermore, the average target coverage (TC) was 98.33% ± 1.16%, dose selectivity (DS) was 0.57 ± 0.10, gradient index (GI) was 2.69 ± 0.30, and homogeneity index (HI) was 1.79 ± 0.09. Compared to the 3D Dense U-Net, CDS-CNN predictions demonstrated a 3.5% improvement in TC, and CDS-CNN's dose prediction yielded better outcomes than the 3D Dense U-Net across all evaluation criteria. The experimental results demonstrated that the proposed CDS-CNN model outperformed other models in predicting GKRS dose distributions, with predictions closely matching the TPS doses.