Sahar Alizada
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Sahar Alizada has a total of 59 co-authors. Here are the top co-authors by number of shared publications:
- Max Wintermark
Department of Radiology, The University of Texas Medical Branch, Galveston
7 shared publications - Hamza Adel Salim
Department of Neuroradiology, MD Anderson Cancer Center, University of Texas, Houston, Texas, USA
5 shared publications - Heba Al Qudah
From the Department of Neuroradiology (R.D., H.A.Q., D.K., S.A., M.W.) and Department of Imaging Physics (P.S., B.A.T., P.H., H.-L.L.), The University of Texas MD Anderson Cancer Center
3 shared publications - Rami W. Eldaya
The University of Texas MD Anderson Cancer Center
3 shared publications Ahmed Msherghi
From the Department of Neuroradiology (P.K., A.N., H.S., S.A., A.M., C.B., M.W., K.S.), Division of Diagnostic Imaging, MD Anderson Cancer Center, Houston, TX and Department of Radiology (M.W.), The University of Texas Medical Branch, Galveston, TX
2 shared publications- Ceylan Altintas Taslicay
Department of Neuroradiology, The University of Texas MD Anderson Cancer Center, Houston, TX
2 shared publications
Latest Publications
- Open Access
- 14 June 2026
- Published by Springer Nature in Acta Neuropathologica Communications
- Vol. 14 (1)
Cystic glioblastoma is associated with alterations of KDR and the Rb pathway: a single-center retrospective analysis
Glioblastoma (GBM) is the most common malignant primary brain tumor in adults. Cystic GBM is a phenotypic subset requiring distinct clinical management. Th
- 14 June 2026
- Published by Springer Nature in Acta Neuropathologica Communications
- Vol. 14 (1)
Open AccessCystic glioblastoma is associated with alterations of KDR and the Rb pathway: a single-center retrospective analysis
Glioblastoma (GBM) is the most common malignant primary brain tumor in adults. Cystic GBM is a phenotypic subset requiring distinct clinical management. Th
- 25 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- p. ajnr.A9372
Evaluating the Utility and Limitations of Machine Learning Tumor Segmentation for Automated Longitudinal Response Assessment in Neuro-Oncology Treatment Response Classification
Machine learning segmentation has emerged in tumor assessment with high performance in volumetric evaluation of brain tumors. It is unclear, however, if such methods can be translated into full end-to-end clinical tools for assessing treatment response. This study aims to assess the ability and current limitations of machine learning of performing fully automated longitudinal Response Assessment in Neuro-Oncology (RANO). An nnU-Net model was trained for segmentation on 4,162 MRIs of pre- and post-treatment brain tumors. A separate dataset of 91 patients was preprocessed, including co-registration for longitudinal lesion tracking. The nnU-Net model was used to segment enhancing tumor on the longitudinal dataset. After excluding patients classified as progression due to T2/FLAIR changes or clinical criteria, 233 pairwise MRI comparisons remained for 60 patients. Using auto-segmented tumor masks, comparisons were classified into Progressive Disease (PD), Partial Response (PR), Stable Disease (SD), and Complete Response (CR) according to the original RANO criteria based on 2D lesion size, measurable criteria, appearance of new lesions, and changes over time. Model results were compared to expert ground truth and evaluated for accuracy in classifying PD and non-PD disease, per class accuracy, and a variety of performance metrics. All discrepant cases were manually reviewed to classify the cause of error. The model was 77.7% (95% CI 70.6%-84.2%) accurate in distinguishing PD and non-PD disease with sensitivity 92.1% (86.2%-97.3%) and specificity 60.4% (43.4%-74.7%). The highest per-class accuracy was PD at 92.1% (86.2%-97.3%), with much lower performance for non-PD categories, worst for PR at 14.3% (0.0%-38.5%). Patient-averaged accuracy was 66.9% (58.2%-75.3%). False positive PD predictions were commonly due to evolving changes around the resection cavity, affecting the morphology and location of enhancement and accounting for 32.8% of SD classifications. Other causes of error included changing morphology but similar volume of enhancement, missed lesions in less typical locations, and radiation necrosis. Automated tumor segmentation is emerging in treatment response assessment, though highly trained segmentation models are still met with limitations in fully automated response assessment due to nuances such as post-surgical enhancement and comparing location and morphology changes between timepoints.
- 25 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- p. ajnr.A9372
Evaluating the Utility and Limitations of Machine Learning Tumor Segmentation for Automated Longitudinal Response Assessment in Neuro-Oncology Treatment Response Classification
Machine learning segmentation has emerged in tumor assessment with high performance in volumetric evaluation of brain tumors. It is unclear, however, if such methods can be translated into full end-to-end clinical tools for assessing treatment response. This study aims to assess the ability and current limitations of machine learning of performing fully automated longitudinal Response Assessment in Neuro-Oncology (RANO). An nnU-Net model was trained for segmentation on 4,162 MRIs of pre- and post-treatment brain tumors. A separate dataset of 91 patients was preprocessed, including co-registration for longitudinal lesion tracking. The nnU-Net model was used to segment enhancing tumor on the longitudinal dataset. After excluding patients classified as progression due to T2/FLAIR changes or clinical criteria, 233 pairwise MRI comparisons remained for 60 patients. Using auto-segmented tumor masks, comparisons were classified into Progressive Disease (PD), Partial Response (PR), Stable Disease (SD), and Complete Response (CR) according to the original RANO criteria based on 2D lesion size, measurable criteria, appearance of new lesions, and changes over time. Model results were compared to expert ground truth and evaluated for accuracy in classifying PD and non-PD disease, per class accuracy, and a variety of performance metrics. All discrepant cases were manually reviewed to classify the cause of error. The model was 77.7% (95% CI 70.6%-84.2%) accurate in distinguishing PD and non-PD disease with sensitivity 92.1% (86.2%-97.3%) and specificity 60.4% (43.4%-74.7%). The highest per-class accuracy was PD at 92.1% (86.2%-97.3%), with much lower performance for non-PD categories, worst for PR at 14.3% (0.0%-38.5%). Patient-averaged accuracy was 66.9% (58.2%-75.3%). False positive PD predictions were commonly due to evolving changes around the resection cavity, affecting the morphology and location of enhancement and accounting for 32.8% of SD classifications. Other causes of error included changing morphology but similar volume of enhancement, missed lesions in less typical locations, and radiation necrosis. Automated tumor segmentation is emerging in treatment response assessment, though highly trained segmentation models are still met with limitations in fully automated response assessment due to nuances such as post-surgical enhancement and comparing location and morphology changes between timepoints.
- 8 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (11) , 2900-2904
Imaging and Clinical Features of Secondary Spinal Cord Lymphoma: A 2-Center Case Series
Secondary CNS lymphoma is an uncommon manifestation of lymphoma with well-established imaging literature describing intracranial findings. Secondary spinal cord involvement by lymphoma is exceedingly rare with limited clinical and imaging data. In this two center case series we present 18 cases of secondary spinal cord lymphoma. We highlight imaging features of secondary spinal cord lymphoma on MRI while establishing its correlation with intracranial disease. We also explore its relationship with systemic disease and pertinent clinical findings.
- 8 April 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (11) , 2900-2904
Imaging and Clinical Features of Secondary Spinal Cord Lymphoma: A 2-Center Case Series
Secondary CNS lymphoma is an uncommon manifestation of lymphoma with well-established imaging literature describing intracranial findings. Secondary spinal cord involvement by lymphoma is exceedingly rare with limited clinical and imaging data. In this two center case series we present 18 cases of secondary spinal cord lymphoma. We highlight imaging features of secondary spinal cord lymphoma on MRI while establishing its correlation with intracranial disease. We also explore its relationship with systemic disease and pertinent clinical findings.
- 11 February 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (8) , ajnr.A9224-2186
Dynamic Susceptibility Contrast MRI using Gadopiclenol for Evaluation of Relative Cerebral Blood Volume at 3T: A Comparison with Gadobutrol
Dynamic susceptibility contrast (DSC) MRI is widely used to assess relative cerebral blood volume (rCBV) in brain tumors using gadolinium-based contrast agents. Gadopiclenol, recently FDA-approved for use at half dose due to its high relaxivity, offers a potential strategy to reduce overall gadolinium exposure. This retrospective study evaluated DSC MRI performance with 0.05 mmol/kg gadopiclenol compared with standard-dose (0.10 mmol/kg) gadobutrol. 20 patients with brain tumors underwent gadopiclenol-based DSC MRI on 3T scanners and were compared with 20 matched controls receiving gadobutrol. Analysis of normal-appearing thalamic, frontal cortical gray matter, and white matter regions showed that gadopiclenol produced DSC time curves with shapes similar to gadobutrol, though with significantly smaller signal decreases and approximately half the peak ΔR2* values, consistent with dose scaling. Importantly, normalized rCBV values did not differ significantly between contrast agents in gray matter regions. These findings support the use of half-dose gadopiclenol for reliable rCBV assessment while reducing gadolinium exposure.
- 11 February 2026
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (8) , ajnr.A9224-2186
Dynamic Susceptibility Contrast MRI using Gadopiclenol for Evaluation of Relative Cerebral Blood Volume at 3T: A Comparison with Gadobutrol
Dynamic susceptibility contrast (DSC) MRI is widely used to assess relative cerebral blood volume (rCBV) in brain tumors using gadolinium-based contrast agents. Gadopiclenol, recently FDA-approved for use at half dose due to its high relaxivity, offers a potential strategy to reduce overall gadolinium exposure. This retrospective study evaluated DSC MRI performance with 0.05 mmol/kg gadopiclenol compared with standard-dose (0.10 mmol/kg) gadobutrol. 20 patients with brain tumors underwent gadopiclenol-based DSC MRI on 3T scanners and were compared with 20 matched controls receiving gadobutrol. Analysis of normal-appearing thalamic, frontal cortical gray matter, and white matter regions showed that gadopiclenol produced DSC time curves with shapes similar to gadobutrol, though with significantly smaller signal decreases and approximately half the peak ΔR2* values, consistent with dose scaling. Importantly, normalized rCBV values did not differ significantly between contrast agents in gray matter regions. These findings support the use of half-dose gadopiclenol for reliable rCBV assessment while reducing gadolinium exposure.
- 22 January 2026
- Published by Elsevier in Clinical Oncology
- Vol. 52, 104049
Whole-Brain Radiation Therapy for Brain Metastases due to Thyroid Cancer: A Systematic Review and Meta-Analysis
- 22 January 2026
- Published by Elsevier in Clinical Oncology
- Vol. 52, 104049
Whole-Brain Radiation Therapy for Brain Metastases due to Thyroid Cancer: A Systematic Review and Meta-Analysis
- 1 October 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 46 (10) , 2116-2124
Differentiation between Nonenhancing Tumor in Glioblastoma and Vasogenic Edema Using Diffusion-Weighted and Dynamic Susceptibility Contrast MR Imaging
BACKGROUND AND PURPOSE: Differentiating non-enhancing glioblastoma from vasogenic edema remains a challenge in neuroimaging. This study evaluates the utility of diffusion-weighted and dynamic susceptibility contrast MR imaging, particularly apparent diffusion coefficient (ADC) and relative cerebral blood volume (rCBV), in distinguishing these entities. MATERIALS AND METHODS: We retrospectively analyzed 111 patients diagnosed with glioblastoma with available pretreatment MRI from January 1, 2022, to December 31, 2023, including 13 patients with 15 solid non-enhancing glioblastoma lesions and 98 patients with perilesional non-enhancing T2 FLAIR hyperintensity surrounding enhancing glioblastomas. Their ADC and rCBV values were compared with those of vasogenic edema surrounding 30 brain metastatic lesions age-and sex-matched to solid non-enhancing glioblastoma lesions. Receiver operating characteristic (ROC) analysis was performed to determine optimal ADC and rCBV thresholds for classification. RESULTS: Solid non-enhancing glioblastomas exhibited significantly lower ADC values and higher rCBV values compared to vasogenic edema (1.08 ± 0.22 × 10-³ vs. 1.74 ± 0.17 × 10-³ mm2/s and 2.4 ± 0.86 vs. 0.30 ± 0.13, respectively; p < 0.001). The optimal ADC and rCBV cutoffs for differentiating solid non-enhancing glioblastoma from vasogenic edema were 1.36 × 10-³ mm2/s and 1.04, achieving excellent sensitivity (0.93-1) and specificity (1). Perilesional non-enhancing T2 FLAIR hyperintensity around enhancing glioblastomas had similar ADC values to vasogenic edema (1.67 ± 0.33 × 10-³ vs. 1.74 ± 0.17 × 10-³ mm2/s, p = 0.32) but significantly higher rCBV (0.60 ± 0.61 vs. 0.30 ± 0.13, p = 0.03). An rCBV cutoff value of 0.42 distinguished perilesional tumor infiltration from pure vasogenic edema with 86% specificity. CONCLUSIONS: Diffusion-and perfusion-weighted MR imaging, particularly ADC and rCBV measurements, provide valuable biomarkers for differentiating non-enhancing glioblastoma from vasogenic edema. The identified threshold values may enhance glioblastoma characterization, improve preoperative differentiation from metastases, and support more precise, image-guided clinical management. ABBREVIATIONS: rCBV = relative cerebral blood volume; IDH = Isocitric Dehydrogenase; KPS = Karnofsky Performance Status Scale; MGMT = O6-methylguanine-DNA methyltransferase; IQR = interquartile ranges; SD = standard deviation.
- 1 October 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 46 (10) , 2116-2124
Differentiation between Nonenhancing Tumor in Glioblastoma and Vasogenic Edema Using Diffusion-Weighted and Dynamic Susceptibility Contrast MR Imaging
BACKGROUND AND PURPOSE: Differentiating non-enhancing glioblastoma from vasogenic edema remains a challenge in neuroimaging. This study evaluates the utility of diffusion-weighted and dynamic susceptibility contrast MR imaging, particularly apparent diffusion coefficient (ADC) and relative cerebral blood volume (rCBV), in distinguishing these entities. MATERIALS AND METHODS: We retrospectively analyzed 111 patients diagnosed with glioblastoma with available pretreatment MRI from January 1, 2022, to December 31, 2023, including 13 patients with 15 solid non-enhancing glioblastoma lesions and 98 patients with perilesional non-enhancing T2 FLAIR hyperintensity surrounding enhancing glioblastomas. Their ADC and rCBV values were compared with those of vasogenic edema surrounding 30 brain metastatic lesions age-and sex-matched to solid non-enhancing glioblastoma lesions. Receiver operating characteristic (ROC) analysis was performed to determine optimal ADC and rCBV thresholds for classification. RESULTS: Solid non-enhancing glioblastomas exhibited significantly lower ADC values and higher rCBV values compared to vasogenic edema (1.08 ± 0.22 × 10-³ vs. 1.74 ± 0.17 × 10-³ mm2/s and 2.4 ± 0.86 vs. 0.30 ± 0.13, respectively; p < 0.001). The optimal ADC and rCBV cutoffs for differentiating solid non-enhancing glioblastoma from vasogenic edema were 1.36 × 10-³ mm2/s and 1.04, achieving excellent sensitivity (0.93-1) and specificity (1). Perilesional non-enhancing T2 FLAIR hyperintensity around enhancing glioblastomas had similar ADC values to vasogenic edema (1.67 ± 0.33 × 10-³ vs. 1.74 ± 0.17 × 10-³ mm2/s, p = 0.32) but significantly higher rCBV (0.60 ± 0.61 vs. 0.30 ± 0.13, p = 0.03). An rCBV cutoff value of 0.42 distinguished perilesional tumor infiltration from pure vasogenic edema with 86% specificity. CONCLUSIONS: Diffusion-and perfusion-weighted MR imaging, particularly ADC and rCBV measurements, provide valuable biomarkers for differentiating non-enhancing glioblastoma from vasogenic edema. The identified threshold values may enhance glioblastoma characterization, improve preoperative differentiation from metastases, and support more precise, image-guided clinical management. ABBREVIATIONS: rCBV = relative cerebral blood volume; IDH = Isocitric Dehydrogenase; KPS = Karnofsky Performance Status Scale; MGMT = O6-methylguanine-DNA methyltransferase; IQR = interquartile ranges; SD = standard deviation.
- 28 August 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (2) , 430-436
Multimodal Imaging and Clinicopathologic Features of Parathyroid Lipoadenoma
Parathyroid lipoadenoma (PLA) is a rare cause of primary hyperparathyroidism and is over 50% adipose tissue, which complicates preoperative localization. We aimed to describe clinical and imaging features of PLA in this case series. We retrospectively reviewed 4 patients with pathologically confirmed PLA and biochemical evidence of primary hyperparathyroidism. All patients exhibited elevated PTH and upper-normal to elevated calcium levels. PLAs demonstrated hyperechogenicity on ultrasonography, fat-equivalent attenuation on CT, and increased uptake on sestamibi SPECT/CT. Surgical excision resulted in significant intraoperative PTH decline and postoperative biochemical cure. Pathology confirmed >50% stromal fat in all cases. In conclusion, PLA presents unique diagnostic challenges due to its imaging characteristics. Awareness of these features and a multimodal imaging approach are key for accurate localization and successful surgical management. ABBREVIATIONS: PLA = parathyroid lipoadenoma, PTH = parathyroid hormone, WHO = World Health Organization, BMI = body mass index, MEN1 = multiple endocrine neoplasia type 1
- 28 August 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 47 (2) , 430-436
Multimodal Imaging and Clinicopathologic Features of Parathyroid Lipoadenoma
Parathyroid lipoadenoma (PLA) is a rare cause of primary hyperparathyroidism and is over 50% adipose tissue, which complicates preoperative localization. We aimed to describe clinical and imaging features of PLA in this case series. We retrospectively reviewed 4 patients with pathologically confirmed PLA and biochemical evidence of primary hyperparathyroidism. All patients exhibited elevated PTH and upper-normal to elevated calcium levels. PLAs demonstrated hyperechogenicity on ultrasonography, fat-equivalent attenuation on CT, and increased uptake on sestamibi SPECT/CT. Surgical excision resulted in significant intraoperative PTH decline and postoperative biochemical cure. Pathology confirmed >50% stromal fat in all cases. In conclusion, PLA presents unique diagnostic challenges due to its imaging characteristics. Awareness of these features and a multimodal imaging approach are key for accurate localization and successful surgical management. ABBREVIATIONS: PLA = parathyroid lipoadenoma, PTH = parathyroid hormone, WHO = World Health Organization, BMI = body mass index, MEN1 = multiple endocrine neoplasia type 1
- 12 June 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 46 (12) , 2697-2704
Clinical and Imaging Features of Primary Intradural Extramedullary Ewing Sarcoma
Primary intradural extramedullary Ewing sarcoma (PIEES) is a very rare and aggressive tumor with poorly recognized radiological features. In this case series, we present our institutional experience with PIEES with emphasis on imaging and temporal evolution of the disease. In total, 8 patients with pathological diagnosis of PIEES were included in this study. All patients were symptomatic at time of presentation. MRI findings demonstrated T1 isointensity and T2 slightly hyperintensity relative to skeletal muscle, T2 hypointensity relative to CSF, and homogeneously enhancement in all cases with spinal cord edema/cord compression or cauda equina nerve roots compression. Tumor recurrence occurred in six patients (6/8, 75%). None of the patients developed recurrence outside the CNS. Four patients (50%) were deceased at the time of our study. In conclusion, PIEES is an extremely rare malignancy with poor outcomes. This case series highlights imaging features of PIEES and its temporal evolution including patterns of recurrence while reinforcing the importance of distinguishing PIEES from benign intradural lesions. ABBREVIATIONS: PIEES = Primary intradural extramedullary Ewing sarcoma, IQR = interquartile range, CI = confidence interval.
- 12 June 2025
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 46 (12) , 2697-2704
Clinical and Imaging Features of Primary Intradural Extramedullary Ewing Sarcoma
Primary intradural extramedullary Ewing sarcoma (PIEES) is a very rare and aggressive tumor with poorly recognized radiological features. In this case series, we present our institutional experience with PIEES with emphasis on imaging and temporal evolution of the disease. In total, 8 patients with pathological diagnosis of PIEES were included in this study. All patients were symptomatic at time of presentation. MRI findings demonstrated T1 isointensity and T2 slightly hyperintensity relative to skeletal muscle, T2 hypointensity relative to CSF, and homogeneously enhancement in all cases with spinal cord edema/cord compression or cauda equina nerve roots compression. Tumor recurrence occurred in six patients (6/8, 75%). None of the patients developed recurrence outside the CNS. Four patients (50%) were deceased at the time of our study. In conclusion, PIEES is an extremely rare malignancy with poor outcomes. This case series highlights imaging features of PIEES and its temporal evolution including patterns of recurrence while reinforcing the importance of distinguishing PIEES from benign intradural lesions. ABBREVIATIONS: PIEES = Primary intradural extramedullary Ewing sarcoma, IQR = interquartile range, CI = confidence interval.
- 14 July 2024
- Published by Elsevier in Clinical Radiology
- Vol. 79 (10) , 722-735
Under pressure: a head-to-toe review of vascular compression syndromes
- 14 July 2024
- Published by Elsevier in Clinical Radiology
- Vol. 79 (10) , 722-735
Under pressure: a head-to-toe review of vascular compression syndromes