A Comparison between the Diagnostic Accuracy of Frozen Section and Permanent Section Analyses in Central Nervous System

Brain tumor grading diagnosis using transfer learning based on optical coherence tomography

In neurosurgery, accurately identifying brain tumor tissue is vital for reducing recurrence. Current imaging techniques have limitations, prompting the exploration of alternative methods. This study validated a binary hierarchical classification of brain tissues: normal tissue, primary central nervous system lymphoma (PCNSL), high-grade glioma (HGG), and low-grade glioma (LGG) using transfer learning. Tumor specimens were measured with optical coherence tomography (OCT), and a MobileNetV2 pre-trained model was employed for classification. Surgeons could optimize predictions based on experience. The model showed robust classification and promising clinical value. A dynamic t-SNE visualized its performance, offering a new approach to neurosurgical decision-making regarding brain tumors.

Confounders of intraoperative frozen section pathology during glioma surgery

Although frozen section pathology (FSP) is commonly performed during surgery for glioma-suspicious lesions, confounders of accuracy are largely unknown. FSP and final diagnosis were compared in 398 surgeries for glioma-suspicious lesions. Diagnostic accuracy, risk factors for diagnostic shift from neoplastic to non-neoplastic tissue and vice versa according to the final diagnosis, and the impact on intraoperative and postoperative decision-making were analyzed. Diagnostic shift occurred in 70 cases (18%), and sensitivity, specificity, and the positive (PPV) and negative (NPV) predictive value of FSP were 82.5%, 77.8%, 99.4%, and 9.3%, respectively. No correlations between shift and patients' age and sex, sample fluorescence or volume, tumor location, correct information on the pathology form, final high- or low-grade histology, or molecular alterations were found (p > .05, each). Shift was more common after irradiation (25% vs 15%; p = .025) or chemotherapy (26% vs 15%; p = .022) than in treatment naïve cases and correlated with the type of surgery (p = .002). FSP altered intraoperative decision-making in 25 cases (6%). Postoperative shift led to repeated surgery in 12 patients (3%). In 45 cases, in which FSP and final diagnosis based on the same tissue, shift occurred in only 5 patients (11%), and sensitivity, specificity, PPV, and NPV for FSP were 77.4%, 78.6%, 88.9%, and 61.1%, respectively. No correlations between diagnostic shift and any of the analyzed variables were found (p > .05, each). Although accuracy of FSP during glioma surgery is sufficient, moderate NPV should be considered during intraoperative decision-making. While confounders are sparse, accuracy might be increased by repeated sampling. Diagnostic shift rarely alters postoperative treatment strategy.

Developing and validating a deep learning and radiomic model for glioma grading using multiplanar reconstructed magnetic resonance contrast-enhanced T1-weighted imaging: a robust, multi-institutional study

Background

Although surgical pathology or biopsy are considered the gold standard for glioma grading, these procedures have limitations. This study set out to evaluate and validate the predictive performance of a deep learning radiomics model based on contrast-enhanced T1-weighted multiplanar reconstruction images for grading gliomas.

Methods

Patients from three institutions who diagnosed with gliomas by surgical specimen and multiplanar reconstructed (MPR) images were enrolled in this study. The training cohort included 101 patients from institution 1, including 43 high-grade glioma (HGG) patients and 58 low-grade glioma (LGG) patients, while the test cohorts consisted of 50 patients from institutions 2 and 3 (25 HGG patients, 25 LGG patients). We then extracted radiomics features and deep learning features using six pretrained models from the MPR images. The Spearman correlation test and the recursive elimination feature selection method were used to reduce the redundancy and select most predictive features. Subsequently, three classifiers were used to construct classification models. The performance of the grading models was evaluated using the area under the receiver operating curve, sensitivity, specificity, accuracy, precision, and negative predictive value. Finally, the prediction performances of the test cohort were compared to determine the optimal classification model.

Results

For the training cohort, 62% (13 out of 21) of the classification models constructed with MPR images from multiple planes outperformed those constructed with single-plane MPR images, and 61% (11 out of 18) of classification models constructed with both radiomics features and deep learning features had higher area under the curve (AUC) values than those constructed with only radiomics or deep learning features. The optimal model was a random forest model that combined radiomic features and VGG16 deep learning features derived from MPR images, which achieved AUC of 0.847 in the training cohort and 0.898 in the test cohort. In the test cohort, the sensitivity, specificity, and accuracy of the optimal model were 0.840, 0.760, and 0.800, respectively.

Conclusions

Multiplanar CE-T1W MPR imaging features are more effective than features from single planes when differentiating HGG and LGG. The combination of deep learning features and radiomics features can effectively grade glioma and assist clinical decision-making.

Evaluation Study of Intraoperative Cytology Smear and Frozen Section of Glioma

Objective:

Glioma is the commonest primary malignant brain tumour. Diagnosis is made based on cytology smear, frozen section and histopathological examination. Intraoperative pathological diagnosis using either cytology smear, frozen section or combination of both, plays a crucial role in patient’s future management and prognosis. This study aims to determine the accuracy of cytology smear and frozen section in glioma, and to compare the difference between both techniques.

Methods:

A cross-sectional study was conducted involving 22 cases of glioma diagnosed intraoperatively from January 2013 until August 2019 in Hospital Universiti Sains Malaysia. The selected tissues were processed for cytology smear and frozen section. The remaining tissues were proceeded for paraffin section. The diagnosis was categorized as either low-grade or high-grade glioma based on cellularity, nuclear pleomorphism, mitotic count, microvascular proliferation and necrosis. The sensitivity and specificity of frozen section and cytology smears were determined based on paraffin section being as the gold standard. The accuracy of both techniques was compared using statistical analysis.

Results:

The overall sensitivity and specificity of cytology smear were 100% and 76.9%, respectively. Meanwhile, the sensitivity and specificity of frozen section were 100% and 84.6%. There was no significant difference in diagnostic accuracy between cytology smear and frozen section in glioma (p>0.05).

Conclusion:

Cytology smears provides an alternative method for frozen section due to good cellularity and morphology on smear. Cytology smear is rapid, inexpensive, small amount of tissue requirement and less technical demand. This finding may benefit to the hospital or treatment centres where frozen section facility is unavailable.

Diagnostic Accuracy of Cytology Smear and Frozen Section in Glioma

Glioma is the commonest primary intracranial tumour and it has been the most predominant tumour in many studies. It accounts for 24.7% of all primary brain tumour and 74.6% of malignant brain tumour. Intraoperative diagnosis plays a crucial role in determining the patient management. Frozen section has been the established technique in providing rapid and accurate intraoperative diagnosis. However due to some disadvantages like ice crystal artefact, high expenditure and requirement of skilled technician, there is increase usage of cytology smear either replacing or supplementing frozen section technique. The aim of this review is to determine the diagnostic accuracy of cytology smear and frozen section in glioma and to see whether there is significant difference between those techniques. The overall diagnostic accuracy for frozen section in glioma ranging from 78.4% to 95% while for cytology smear, the diagnostic accuracy ranging from 50% to 100%. Based on certain literatures, no statistically difference was observed in diagnostic accuracy of cytology smear and frozen section. Thus, cytology smear provides an alternative method in establishing intraoperative diagnosis. Both cytology smear and frozen section are complimentary to each other. It is recommended to use both techniques to improve the diagnostic accuracy in addition with adequate knowledge, clinical history, neuroimaging and intraoperative findings.