Automated histologic diagnosis of CNS tumors with machine learning

Algorithm-assisted diagnosis of Hirschsprung's disease - evaluation of robustness and comparative image analysis on data from various labs and slide scanners

BACKGROUND

Differences in the preparation, staining and scanning of digital pathology slides create significant pre-analytic variability. Algorithm-assisted tools must be able to contend with this variability in order to be applicable in clinical practice. In a previous study, a decision support algorithm was developed to assist in the diagnosis of Hirschsprung's disease. In the current study, we tested the robustness of this algorithm while assessing for pre-analytic factors which may affect its performance.

METHODS

The decision support algorithm was used on digital pathology slides obtained from four different medical centers (A-D) and scanned by three different scanner models (by Philips, Hamamatsu and 3DHISTECH). A total of 192 cases and 1782 slides were used in this study. RGB histograms were constructed to compare images from the various medical centers and scanner models and highlight the differences in color and contrast.

RESULTS

The algorithm was able to correctly identify ganglion cells in 99.2% of cases, from all medical centers (All scanned by the Philips slide scanner) as well as 95.5% and 100% of the slides scanned by the 3DHISTECH and Hamamatsu brand slide scanners, respectively. The total error rate for center D was lower than the other medical centers (3.9% vs 7.1%, 10.8% and 6% for centers A-C, respectively), the vast majority of errors being false positives (3.45% vs 0.45% false negatives). The other medical centers showed a higher rate of false negatives in relation to false positives (6.81% vs 0.29%, 9.8% vs 1.2% and 5.37% vs 0.63% for centers A-C, respectively). The total error rates for the Philips, Hamamatsu and 3DHISTECH brand scanners were 3.9%, 3.2% and 9.8%, respectively. RGB histograms demonstrated significant differences in pixel value distribution between the four medical centers, as well as between the 3DHISTECH brand scanner when compared to the Philips and Hamamatsu brand scanners.

CONCLUSIONS

The results reported in this paper suggest that the algorithm-based decision support system has sufficient robustness to be applicable for clinical practice. In addition, the novel method used in its development - Hierarchial-Contexual Analysis (HCA) may be applicable to the development of algorithm-assisted tools in other diseases, for which available datasets are limited. Validation of any given algorithm-assisted support system should nonetheless include data from as many medical centers and scanner models as possible.

Deep Learning-Assisted Segmentation and Classification of Brain Tumor Types on Magnetic Resonance and Surgical Microscope Images

OBJECTIVE

The primary aim of this research was to harness the capabilities of deep learning to enhance neurosurgical procedures, focusing on accurate tumor boundary delineation and classification. Through advanced diagnostic tools, we aimed to offer surgeons a more insightful perspective during surgeries, improving surgical outcomes and patient care.

METHODS

The study deployed the Mask R-convolutional neural network (CNN) architecture, leveraging its sophisticated features to process and analyze data from surgical microscope videos and preoperative magnetic resonance images. Resnet101 and Resnet50 backbone networks are used in the Mask R-CNN method, and experimental results are given. We subsequently tested its performance across various metrics, such as accuracy, precision, recall, dice coefficient (DICE), and Jaccard index. Deep learning models were trained from magnetic resonance imaging and surgical microscope images, and the classification result obtained for each patient was combined with the weighted average.

RESULTS

The algorithm exhibited remarkable capabilities in distinguishing among meningiomas, metastases, and high-grade glial tumors. Specifically, for the Mask R-CNN Resnet 101 architecture, precision, recall, DICE, and Jaccard index values were recorded as 96%, 93%, 91%, and 84%, respectively. Conversely, for the Mask R-CNN Resnet 50 architecture, these values stood at 94%, 89%, 89%, and 82%. Additionally, the model achieved an impressive DICE score range of 94%-95% and an accuracy of 98% in pathology estimation.

CONCLUSIONS

As illustrated in our study, the confluence of deep learning with neurosurgical procedures marks a transformative phase in medical science. The results are promising but underscore diverse data sets' significance for training and refining these deep learning models.

Unclassifiable CNS tumors in DNA methylation-based classification: clinical challenges and prognostic impact

DNA methylation analysis has become a powerful tool in neuropathology. Although DNA methylation-based classification usually shows high accuracy, certain samples cannot be classified and remain clinically challenging. We aimed to gain insight into these cases from a clinical perspective. To address, central nervous system (CNS) tumors were subjected to DNA methylation profiling and classified according to their calibrated score using the DKFZ brain tumor classifier (V11.4) as "≥ 0.84" (score ≥ 0.84), "0.3-0.84" (score 0.3-0.84), or "< 0.3" (score < 0.3). Histopathology, patient characteristics, DNA input amount, and tumor purity were correlated. Clinical outcome parameters were time to treatment decision, progression-free, and overall survival. In 1481 patients, the classifier identified 69 (4.6%) tumors with an unreliable score as "< 0.3". Younger age (P < 0.01) and lower tumor purity (P < 0.01) compromised accurate classification. A clinical impact was demonstrated as unclassifiable cases ("< 0.3") had a longer time to treatment decision (P < 0.0001). In a subset of glioblastomas, these cases experienced an increased time to adjuvant treatment start (P < 0.001) and unfavorable survival (P < 0.025). Although DNA methylation profiling adds an important contribution to CNS tumor diagnostics, clinicians should be aware of a potentially longer time to treatment initiation, especially in malignant brain tumors.

Computer Vision in Digital Neuropathology

Digital pathology has revolutionized the field of neuropathology, enabling rapid and precise analysis of tissue samples. As neuropathologists and neurosurgeons increasingly embrace digital platforms, computer vision techniques and computerized quantitative analyses (i.e., pathomics) play a pivotal role in automating and enhancing diagnostic processes, as well as in unlocking new insights from neuropathological images. This chapter explores the role of computer vision in bridging the gap between traditional and digital neuropathology, providing a comprehensive overview of computer vision applications in neuropathology, covering image preprocessing, and decision support for clinical tasks, such as early detection of neurological disorders, classification of brain tumors, and quantitative assessment of tissue biomarkers. In addition, we further delve into the challenges and opportunities presented by large-scale image datasets and the integration of digital pathology into neurosurgical practice.

[Artificial intelligence in pathological anatomy]

The review presents key concepts and global developments in the field of artificial intelligence used in pathological anatomy. The work examines two types of artificial intelligence (AI): weak and strong ones. A review of experimental algorithms using both deep machine learning and computer vision technologies to work with WSI images of preparations, diagnose and make a prognosis for various malignant neoplasms is carried out. It has been established that weak artificial intelligence at this stage of development of computer (digital) pathological anatomy shows significantly better results in speeding up and refining diagnostic procedures than strong artificial intelligence having signs of general intelligence. The article also discusses three options for the further development of AI assistants for pathologists based on the technologies of large language models (strong AI) ChatGPT (PathAsst), Flan-PaLM2 and LIMA. As a result of the analysis of the literature, key problems in the field were identified: the equipment of pathology institutions, the lack of experts in training neural networks, the lack of strict criteria for the clinical viability of AI diagnostic technologies.

Neurosurgical Management of Cerebrospinal Tumors in the Era of Artificial Intelligence : A Scoping Review

Central nervous system tumors are identified as tumors of the brain and spinal cord. The associated morbidity and mortality of cerebrospinal tumors are disproportionately high compared to other malignancies. While minimally invasive techniques have initiated a revolution in neurosurgery, artificial intelligence (AI) is expediting it. Our study aims to analyze AI's role in the neurosurgical management of cerebrospinal tumors. We conducted a scoping review using the Arksey and O'Malley framework. Upon screening, data extraction and analysis were focused on exploring all potential implications of AI, classification of these implications in the management of cerebrospinal tumors. AI has enhanced the precision of diagnosis of these tumors, enables surgeons to excise the tumor margins completely, thereby reducing the risk of recurrence, and helps to make a more accurate prediction of the patient's prognosis than the conventional methods. AI also offers real-time training to neurosurgeons using virtual and 3D simulation, thereby increasing their confidence and skills during procedures. In addition, robotics is integrated into neurosurgery and identified to increase patient outcomes by making surgery less invasive. AI, including machine learning, is rigorously considered for its applications in the neurosurgical management of cerebrospinal tumors. This field requires further research focused on areas clinically essential in improving the outcome that is also economically feasible for clinical use. The authors suggest that data analysts and neurosurgeons collaborate to explore the full potential of AI.

Epigenetic profiling reveals a strong association between lack of 5-ALA fluorescence and EGFR amplification in IDH-wildtype glioblastoma

Background

5-aminolevulinic acid (5-ALA) fluorescence-guided resection increases the percentage of complete CNS tumor resections and improves the progression-free survival of IDH-wildtype glioblastoma patients. A small subset of IDH-wildtype glioblastoma shows no 5-ALA fluorescence. An explanation for these cases is missing. In this study, we used DNA methylation profiling to further characterize non-fluorescent glioblastomas.

Methods

Patients with newly diagnosed and recurrent IDH-wildtype glioblastoma that underwent surgery were analyzed. The intensity of intraoperative 5-ALA fluorescence was categorized as non-visible or visible. DNA was extracted from tumors and genome-wide DNA methylation patterns were analyzed using Illumina EPIC (850k) arrays. Furthermore, 5-ALA intensity was measured by flow cytometry on human gliomasphere lines (BT112 and BT145).

Results

Of 74 included patients, 12 (16.2%) patients had a non-fluorescent glioblastoma, which were compared to 62 glioblastomas with 5-ALA fluorescence. Clinical characteristics were equally distributed between both groups. We did not find significant differences between DNA methylation subclasses and 5-ALA fluorescence (P = .24). The distribution of cells of the tumor microenvironment was not significantly different between the non-fluorescent and fluorescent tumors. Copy number variations in EGFR and simultaneous EGFRvIII expression were strongly associated with 5-ALA fluorescence since all non-fluorescent glioblastomas were EGFR-amplified (P < .01). This finding was also demonstrated in recurrent tumors. Similarly, EGFR-amplified glioblastoma cell lines showed no 5-ALA fluorescence after 24 h of incubation.

Conclusions

Our study demonstrates an association between non-fluorescent IDH-wildtype glioblastomas and EGFR gene amplification which should be taken into consideration for recurrent surgery and future studies investigating EGFR-amplified gliomas.

Histopathological Images Analysis and Predictive Modeling Implemented in Digital Pathology-Current Affairs and Perspectives

In modern clinical practice, digital pathology has an essential role, being a technological necessity for the activity in the pathological anatomy laboratories. The development of information technology has majorly facilitated the management of digital images and their sharing for clinical use; the methods to analyze digital histopathological images, based on artificial intelligence techniques and specific models, quantify the required information with significantly higher consistency and precision compared to that provided by optical microscopy. In parallel, the unprecedented advances in machine learning facilitate, through the synergy of artificial intelligence and digital pathology, the possibility of diagnosis based on image analysis, previously limited only to certain specialties. Therefore, the integration of digital images into the study of pathology, combined with advanced algorithms and computer-assisted diagnostic techniques, extends the boundaries of the pathologist's vision beyond the microscopic image and allows the specialist to use and integrate his knowledge and experience adequately. We conducted a search in PubMed on the topic of digital pathology and its applications, to quantify the current state of knowledge. We found that computer-aided image analysis has a superior potential to identify, extract and quantify features in more detail compared to the human pathologist's evaluating possibilities; it performs tasks that exceed its manual capacity, and can produce new diagnostic algorithms and prediction models applicable in translational research that are able to identify new characteristics of diseases based on changes at the cellular and molecular level.

Intraoperative cytological diagnosis of brain tumours: A preliminary study using a deep learning model

BACKGROUND

Intraoperative pathological diagnosis of central nervous system (CNS) tumours is essential to planning patient management in neuro-oncology. Frozen section slides and cytological preparations provide architectural and cellular information that is analysed by pathologists to reach an intraoperative diagnosis. Progress in the fields of artificial intelligence and machine learning means that AI systems have significant potential for the provision of highly accurate real-time diagnosis in cytopathology.

OBJECTIVE

To investigate the efficiency of machine-learning models in the intraoperative cytological diagnosis of CNS tumours.

MATERIALS AND METHODS

We trained a deep neural network to classify biopsy material for intraoperative tissue diagnosis of four major brain lesions. Overall, 205 medical images were obtained from squash smear slides of histologically correlated cases, with 18 high-grade and 11 low-grade gliomas, 17 metastatic carcinomas, and 9 non-neoplastic pathological brain tissue samples. The neural network model was trained and evaluated using 5-fold cross-validation.

RESULTS

The model achieved 95% and 97% diagnostic accuracy in the patch-level classification and patient-level classification tasks, respectively.

CONCLUSIONS

We conclude that deep learning-based classification of cytological preparations may be a promising complementary method for the rapid and accurate intraoperative diagnosis of CNS tumours.

Fully Automated MRI Segmentation and Volumetric Measurement of Intracranial Meningioma Using Deep Learning

BACKGROUND

Accurate and rapid measurement of the MRI volume of meningiomas is essential in clinical practice to determine the growth rate of the tumor. Imperfect automation and disappointing performance for small meningiomas of previous automated volumetric tools limit their use in routine clinical practice.

PURPOSE

To develop and validate a computational model for fully automated meningioma segmentation and volume measurement on contrast-enhanced MRI scans using deep learning.

STUDY TYPE

Retrospective.

POPULATION

A total of 659 intracranial meningioma patients (median age, 59.0 years; interquartile range: 53.0-66.0 years) including 554 women and 105 men.

FIELD STRENGTH/SEQUENCE

The 1.0 T, 1.5 T, and 3.0 T; three-dimensional, T₁ -weighted gradient-echo imaging with contrast enhancement.

ASSESSMENT

The tumors were manually segmented by two neurosurgeons, H.K. and C.-K.P., with 10 and 26 years of clinical experience, respectively, for use as the ground truth. Deep learning models based on U-Net and nnU-Net were trained using 459 subjects and tested for 100 patients from a single institution (internal validation set [IVS]) and 100 patients from other 24 institutions (external validation set [EVS]), respectively. The performance of each model was evaluated with the Sørensen-Dice similarity coefficient (DSC) compared with the ground truth.

STATISTICAL TESTS

According to the normality of the data distribution verified by the Shapiro-Wilk test, variables with three or more categories were compared by the Kruskal-Wallis test with Dunn's post hoc analysis.

RESULTS

A two-dimensional (2D) nnU-Net showed the highest median DSCs of 0.922 and 0.893 for the IVS and EVS, respectively. The nnU-Nets achieved superior performance in meningioma segmentation than the U-Nets. The DSCs of the 2D nnU-Net for small meningiomas less than 1 cm³ were 0.769 and 0.780 with the IVS and EVS, respectively.

DATA CONCLUSION

A fully automated and accurate volumetric measurement tool for meningioma with clinically applicable performance for small meningioma using nnU-Net was developed.

EVIDENCE LEVEL

3 TECHNICAL EFFICACY: Stage 2.

Current Applications of Machine Learning for Spinal Cord Tumors

Spinal cord tumors constitute a diverse group of rare neoplasms associated with significant mortality and morbidity that pose unique clinical and surgical challenges. Diagnostic accuracy and outcome prediction are critical for informed decision making and can promote personalized medicine and facilitate optimal patient management. Machine learning has the ability to analyze and combine vast amounts of data, allowing the identification of patterns and the establishment of clinical associations, which can ultimately enhance patient care. Although artificial intelligence techniques have been explored in other areas of spine surgery, such as spinal deformity surgery, precise machine learning models for spinal tumors are lagging behind. Current applications of machine learning in spinal cord tumors include algorithms that improve diagnostic precision by predicting genetic, molecular, and histopathological profiles. Furthermore, artificial intelligence-based systems can assist surgeons with preoperative planning and surgical resection, potentially reducing the risk of recurrence and consequently improving clinical outcomes. Machine learning algorithms promote personalized medicine by enabling prognostication and risk stratification based on accurate predictions of treatment response, survival, and postoperative complications. Despite their promising potential, machine learning models require extensive validation processes and quality assessments to ensure safe and effective translation to clinical practice.

Stimulated Raman Histology for Rapid Intra-Operative Diagnosis of Sinonasal and Skull Base Tumors

OBJECTIVE

Intra-operative stimulated Raman histology (SRH) is a novel technology that uses laser spectroscopy and color-matching algorithms to create images similar to the formalin-fixed paraffin-embedded (FFPE) section. We aim to assess the accuracy of SRH in a novel range of sinonasal and skull base tumors.

METHODS

Select patients undergoing sinonasal and skull base surgery using the Invenio Imaging™ Nio™ Laser Imaging SRH system between June 2020 and September 2021 were assessed. The SRH images were reviewed for pathologic features similar to frozen section (FS) and FFPE. Time taken for results and diagnostic concordance was assessed.

RESULTS

Sixty-seven SRH images from 7 tumor types in 12 patients were assessed. Pathologies included squamous cell carcinoma, rhabdomyosarcoma, inverted papilloma, adenoid cystic carcinoma, SMARCB1-deficient sinonasal carcinoma, mucosal melanoma, metastatic colonic adenocarcinoma, and meningioma. Tumor was identified in 100% of lesional specimens, with characteristic diagnostic features readily appreciable on SRH. Median time for diagnosis was significantly faster for SRH (4.3 min) versus FS (44.5 min; p = <.0001). Where SRH sample site matched precisely to FS (n = 32/67, 47.8%), the same diagnosis was confirmed in 93.8%. Sensitivity, specificity, precision, and overall accuracy of SRH were 93.3%, 94.1%, 93.8%, and 93.3%, respectively. Near-perfect concordance was seen between SRH and FS (Cohen's kappa [κ] = 0.89).

CONCLUSION

Stimulated Raman histology can rapidly produce images similar to FFPE H&E in sinonasal and skull base tumors. This technology has the potential to act as an adjunct or alternative to standard FS.

LEVEL OF EVIDENCE

4 Laryngoscope, 132:2142-2147, 2022.

A random forest algorithm predicting model combining intraoperative frozen section analysis and clinical features guides surgical strategy for peripheral solitary pulmonary nodules

Background

Intraoperative frozen section (FS) analysis has been used to guide the extent of resection in patients with solitary pulmonary nodules (SPNs), but its accuracy varies greatly among different hospitals. Artificial intelligence (AI) and multidimensional data technology are developing rapidly these years, meanwhile, surgeons need better methods to guide the surgical strategy of SPNs. We established predicting models combining FS results with multidimensional perioperative clinical features using logistic regression analysis and the random forest (RF) algorithm to get more accurate extent of SPN resection.

Methods

Patients with peripheral SPNs who underwent FS-guided surgical resection at the Shanghai Chest Hospital (January 2017-December 2018) were retrospectively examined (N=3,089). The accuracy of intraoperative FS-guided resection extent was analyzed and used as Model 1. The clinical features (sex, age, CT features, tumor markers, smoking history, lesion size and nodule location) of patients were collected, and Models 2 and 3 were established using logistic regression and RF algorithms to combine the FS with clinical features. We confirmed the performance of these models in an external validation cohort of 117 patients from Hwa Mei Hospital, University of Chinese Academy of Science (Ningbo No. 2 Hospital). We compared the effectiveness in classifying low/high-risk groups of SPN among them.

Results

The accuracy of FS analysis was 61.3%. Model 3 exhibited the best diagnostic accuracy and had an area under the curve of 0.903 in n the internal validation cohort and 0.919 in the external validation cohort. The calibration plots and net reclassification index (NRI) of Model 3 also exhibited significantly better performance than the other models. Improved diagnostic accuracy was observed in in both internal and external validation cohort.

Conclusions

Using an RF algorithm, clinical characteristics can be combined with intraoperative FS analysis to significantly improve intraoperative judgment accuracy for low- and high-risk tumors, and may serve as a reliable complementary method when FS evaluation is equivocal, improving the accuracy of the extent of surgical resection.

Deep Interpretable Classification and Weakly-Supervised Segmentation of Histology Images via Max-Min Uncertainty

Weakly-supervised learning (WSL) has recently triggered substantial interest as it mitigates the lack of pixel-wise annotations. Given global image labels, WSL methods yield pixel-level predictions (segmentations), which enable to interpret class predictions. Despite their recent success, mostly with natural images, such methods can face important challenges when the foreground and background regions have similar visual cues, yielding high false-positive rates in segmentations, as is the case in challenging histology images. WSL training is commonly driven by standard classification losses, which implicitly maximize model confidence, and locate the discriminative regions linked to classification decisions. Therefore, they lack mechanisms for modeling explicitly non-discriminative regions and reducing false-positive rates. We propose novel regularization terms, which enable the model to seek both non-discriminative and discriminative regions, while discouraging unbalanced segmentations. We introduce high uncertainty as a criterion to localize non-discriminative regions that do not affect classifier decision, and describe it with original Kullback-Leibler (KL) divergence losses evaluating the deviation of posterior predictions from the uniform distribution. Our KL terms encourage high uncertainty of the model when the latter inputs the latent non-discriminative regions. Our loss integrates: (i) a cross-entropy seeking a foreground, where model confidence about class prediction is high; (ii) a KL regularizer seeking a background, where model uncertainty is high; and (iii) log-barrier terms discouraging unbalanced segmentations. Comprehensive experiments and ablation studies over the public GlaS colon cancer data and a Camelyon16 patch-based benchmark for breast cancer show substantial improvements over state-of-the-art WSL methods, and confirm the effect of our new regularizers (our code is publicly available at https://github.com/sbelharbi/deep-wsl-histo-min-max-uncertainty).

Recent advances in the use of stimulated Raman scattering in histopathology

Stimulated Raman histopathology (SRH) utilises the intrinsic vibrational properties of lipids, proteins and nucleic acids to generate contrast providing rapid image acquisition that allows visualisation of histopathological features. It is currently being trialled in the intraoperative setting, where the ability to image unprocessed samples rapidly and with high resolution offers several potential advantages over the use of conventional haematoxylin and eosin stained images. Here we review recent advances in the field including new updates in instrumentation and computer aided diagnosis. We also discuss how other non-linear modalities can be used to provide additional diagnostic contrast which together pave the way for enhanced histopathology and open up possibilities for in vivo pathology.