102 AI-Based Molecular Classification of Diffuse Gliomas using Rapid, Label-Free Optical Imaging.

INTRODUCTION: Molecular classification has transformed the management of brain tumors by enabling more accurate prognostication and personalized treatment. Access to timely molecular diagnostic testing for brain tumor patients is limited, complicating surgical and adjuvant treatment and obstructing clinical trial enrollment. METHODS: By combining stimulated Raman histology (SRH), a rapid, label-free, non-consumptive, optical imaging method, and deep learning-based image classification, we are able to predict the molecular genetic features used by the World Health Organization (WHO) to define the adult-type diffuse glioma taxonomy, including IDH-1/2, 1p19q-codeletion, and ATRX loss. We developed a multimodal deep neural network training strategy that uses both SRH images and large-scale, public diffuse glioma genomic data (i.e. TCGA, CGGA, etc.) in order to achieve optimal molecular classification performance. RESULTS: One institution was used for model training (University of Michigan) and four institutions (NYU, UCSF, Medical University of Vienna, and University Hospital Cologne) were included for patient enrollment in the prospective testing cohort. Using our system, called DeepGlioma, we achieved an average molecular genetic classification accuracy of 93.2% and identified the correct diffuse glioma molecular subgroup with 91.5% accuracy within 2 minutes in the operating room. DeepGlioma outperformed conventional IDH1-R132H immunohistochemistry (94.2% versus 91.4% accuracy) as a first-line molecular diagnostic screening method for diffuse gliomas and can detect canonical and non-canonical IDH mutations. CONCLUSIONS: Our results demonstrate how artificial intelligence and optical histology can be used to provide a rapid and scalable alternative to wet lab methods for the molecular diagnosis of brain tumor patients during surgery.

Rapid, label-free detection of diffuse glioma recurrence using intraoperative stimulated Raman histology and deep neural networks.

BACKGROUND: Detection of glioma recurrence remains a challenge in modern neuro-oncology. Noninvasive radiographic imaging is unable to definitively differentiate true recurrence versus pseudoprogression. Even in biopsied tissue, it can be challenging to differentiate recurrent tumor and treatment effect. We hypothesized that intraoperative stimulated Raman histology (SRH) and deep neural networks can be used to improve the intraoperative detection of glioma recurrence. METHODS: We used fiber laser-based SRH, a label-free, nonconsumptive, high-resolution microscopy method (<60 sec per 1 × 1 mm2) to image a cohort of patients (n = 35) with suspected recurrent gliomas who underwent biopsy or resection. The SRH images were then used to train a convolutional neural network (CNN) and develop an inference algorithm to detect viable recurrent glioma. Following network training, the performance of the CNN was tested for diagnostic accuracy in a retrospective cohort (n = 48). RESULTS: Using patch-level CNN predictions, the inference algorithm returns a single Bernoulli distribution for the probability of tumor recurrence for each surgical specimen or patient. The external SRH validation dataset consisted of 48 patients (recurrent, 30; pseudoprogression, 18), and we achieved a diagnostic accuracy of 95.8%. CONCLUSION: SRH with CNN-based diagnosis can be used to improve the intraoperative detection of glioma recurrence in near-real time. Our results provide insight into how optical imaging and computer vision can be combined to augment conventional diagnostic methods and improve the quality of specimen sampling at glioma recurrence.

Neurosurgical patients admitted via the emergency department initiating comfort care measures: a prospective cohort analysis.

BACKGROUND: Given the serious nature of many neurosurgical pathologies, it is common for hospitalized patients to elect comfort care (CC) over aggressive treatment. Few studies have evaluated the incidence and risk factors of CC trends in patients admitted for neurosurgical emergencies. OBJECTIVES: To analyze all neurosurgical patients admitted to a tertiary care academic referral center via the emergency department (ED) to determine incidence and characteristics of those who initiated CC measures during their initial hospital admission. METHODS: We performed a prospective, cohort analysis of all consecutive adult patients admitted to the neurosurgical service via the ED between October 2018 and May 2019. The primary outcome was the initiation of CC measures during the patient's hospital admission. CC was defined as cessation of life-sustaining measures and a shift in focus to maintaining the comfort and dignity of the patient. RESULTS: Of the 428 patients admitted during the 7-month period, 29 (6.8%) initiated CC measures within 4.0 ± 4.0 days of admission. Patients who entered CC were significantly more likely to have a medical history of cerebrovascular disease (58.6% vs. 33.3%, p = 0.006), dementia (17.2% vs. 1.5%, p = 0.0004), or cancer with metastatic disease (24.1% vs. 7.0%, p = 0.001). Patients with a presenting pathology associated with cerebrovascular disease were significantly more likely to initiate CC (62.1% vs. 35.3, p = 0.04). Patients who underwent emergent surgery were significantly more likely to enter CC compared with those who had elective surgery (80.0% vs. 42.7%, p = 0.02). Only 10 of the 29 (34.5%) patients who initiated CC underwent a neurosurgical operation (p = 0.002). Twenty of the 29 (69.0%) patients died within 0.8 ± 0.8 days after the initiation of CC measures. CONCLUSION: CC measures were initiated in 6.8% of patients admitted to the neurosurgical service via the ED, with the majority of patients entering CC before an operation and presenting with a cerebrovascular pathology.

Near real-time intraoperative brain tumor diagnosis using stimulated Raman histology and deep neural networks.

Intraoperative diagnosis is essential for providing safe and effective care during cancer surgery1. The existing workflow for intraoperative diagnosis based on hematoxylin and eosin staining of processed tissue is time, resource and labor intensive2,3. Moreover, interpretation of intraoperative histologic images is dependent on a contracting, unevenly distributed, pathology workforce4. In the present study, we report a parallel workflow that combines stimulated Raman histology (SRH)5-7, a label-free optical imaging method and deep convolutional neural networks (CNNs) to predict diagnosis at the bedside in near real-time in an automated fashion. Specifically, our CNNs, trained on over 2.5 million SRH images, predict brain tumor diagnosis in the operating room in under 150 s, an order of magnitude faster than conventional techniques (for example, 20-30 min)2. In a multicenter, prospective clinical trial (n = 278), we demonstrated that CNN-based diagnosis of SRH images was noninferior to pathologist-based interpretation of conventional histologic images (overall accuracy, 94.6% versus 93.9%). Our CNNs learned a hierarchy of recognizable histologic feature representations to classify the major histopathologic classes of brain tumors. In addition, we implemented a semantic segmentation method to identify tumor-infiltrated diagnostic regions within SRH images. These results demonstrate how intraoperative cancer diagnosis can be streamlined, creating a complementary pathway for tissue diagnosis that is independent of a traditional pathology laboratory.